XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver XE8802 Sensing Machine Data Acquisition SoC with 16+10 bit ZoomingADC™ Embedded RISC controller and LCD driver General Description The XE8802 includes a high resolution acquisition path with the 16+10 bits ZoomingADC™ and an LCD driver for up to 120 segments. The LCD lines can be used as additional I/Os. Key product Features • • • The embedded core of the XE8802 is an ultra low-power low-voltage microcontroller unit (MCU) with a strict 1 instruction per clock processing and branch speed, allowing for a sustained 4MIPS at 1200µA under 2.4 to 5.5 V. The XE8802 is available with on chip ROM or Multiple-Time-Programmable (MTP) program memory. Applications • • • • • • Portable, battery operated instruments Medical portable instrument Remote control HVAC control Metering Sports watches, wrist instruments Rev 1 January 2006 • • • • • • • Low-power, high resolution ZoomingADC • 16-bits sigma-delta type ADC • 0.5 to 1000 gain with offset cancellation • up to 13 inputs multiplexer, 8 dedicated inputs 4 low power comparators Low-voltage low-power controller operation • • • • 2 or 4 MIPS max, 2.4 to 5.5 V operation 300µA per 1MIPS over full voltage range 7MIPS or 1.2V operation in ROM Hardware MULT up to 22kByte (8 kInstruction) MTP 1032Byte RAM data memory RC and crystal oscillators 5 reset, 22 interrupt, 8 event sources 36 I/O lines, hardware UART and SPI 120 segments LCD driver, can be used as 32 extra I/O lines 100 years MTP Flash retention at 55°C Ordering Information Product Temperature range Memory type Package XE8802MI000 -40°C to 85 °C MTP die XE8802MI035LF -40°C to 85 °C MTP LQFP100 www.semtech.com XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver TABLE OF CONTENTS Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11 Chapter 12 Chapter 13 Chapter 14 Chapter 15 Chapter 16 Chapter 17 Chapter 18 Chapter 19 Chapter 20 Chapter 21 Chapter 22 Chapter 23 © Semtech 2006 XE8802 Overview XE8802 Performance XE8802 CPU XE8802 Memory Low power modes Reset generator Clock generation Interrupt handler Event handler Low power RAM Port A Port B Port D Universal Asynchronous Receiver/Transmitter (UART) Universal Synchronous Receiver/Transmitter (USRT) Serial Peripheral Interface (SPI) Acquisition chain Voltage multiplier LCD driver Counters/PWM The Voltage Level Detector Low Power Comparators XE8802 Dimensions www.semtech.com XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 1. General overview 1.1 Top schematic 1-2 1.2 1.2.1 1.2.2 Pin map LQFP-100 LQFP-80 1-4 1-4 1-6 1.3 Bare die 1-8 1.4 Pin assignment 1-9 © Semtech 2006 www.semtech.com 1-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 1.1 Top schematic The top-level block schematic of the circuit is shown in Figure 1-1. The heart of the circuit consists of the Coolrisc816 CPU core. This core includes an 8x8 multiplier and 16 internal registers. The bus controller generates all control signals for access to all data registers other than the CPU internal registers. The reset block generates the adequate reset signals for the rest of the circuit as a function of the set-up contained in its control registers. Possible reset sources are the power-on-reset (POR), the external pin NRESET, the watchdog (WD), a bus error detected by the bus controller or a programmable pattern on Port A. Different low power modes are implemented. The clock generation and power management block sets up the clock signals and generates internal supplies for different blocks. The clock can be generated from the RC oscillator (this is the start-up condition), the crystal oscillator (XTAL) or an external clock source (given on the XIN pin). The test controller generates all set-up signals for different test modes. In normal operation, it is used as a set of 8 low power data registers. If power consumption is important for the application, the variables that need to be accessed frequently should be stored in these registers rather than in the RAM. The IRQ handler routes the interrupt signals of the different peripherals to the IRQ inputs of the CPU core. It allows masking of the interrupt sources and it flags which interrupt source is active. Events are generally used to restart the processor after a HALT period without jumping to a specified address, i.e. the program execution resumes with the instruction following the HALT instruction. The EVN handler routes the event signals of the different peripherals to the EVN inputs of the CPU core. It allows masking of the interrupt sources and it flags which interrupt source is active. The Port B is an 8-bit parallel IO port with analog capabilities. The URST, UART, PWM and CMPD block also make use of this port. The instruction memory is a 22-bit wide flash or ROM memory depending on the circuit version. In case of the ROM version, the VPP and NFASTREAD pins are not used. ROM versions have 8k instruction memory. Flash versions have 8k or 4k instruction memories depending on the selected operation speed. The data memory on this product is a 1024 byte SRAM. The Acquisition Chain is a high-resolution acquisition path with the 16+10 bits ZoomingADC™. The VMULT (voltage multiplier) powers a part of the Acquisition Chain. The SPI is a serial interface with a master or slave configuration capability. When unused, the 4 SPI pads can be used as 4-bit wide general-purpose I/O port. The port A is an 8 bit parallel input port. It can also generate interrupts, events or a reset. It can be used to input external clocks for the timer/counter/PWM block. The Port D1 and the Port D2 are two general-purpose 8 bit parallel I/O ports. The LCD driver can support a direct drive display (up to 32 segments), or multiplex 1/2, 1/3, 1/4 displays (up to 120 segments). The driver contains an on chip low-power voltage generation device VGEN. The LCD lines can be used as additional I/O pins. The USRT (universal synchronous receiver/transmitter) contains some simple hardware functions in order to simplify the software implementation of a synchronous serial link. © Semtech 2006 www.semtech.com 1-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver NFASTREAD DATA MEMORY INSTRUCTION MEMORY VPP CPU COOLRISC816 VBAT VSS 8 X 8 MULTIPLIER B U S C O N T R O L L E R address control PORT A PA(7:0) datain dataout PORT D1 PD1(7:0) 16 CPU REGISTERS NRESET RESET BLOCK WD XIN XOUT VREG TEST reset control PORT D2 POR PD2(7:0) CLOCK GENERATION/ POWER MANAGEMENT RC XTAL VREG clocks LCD_IO(31:0) LCD Driver VGEN test control LCD_COM(1:0) TEST CONTROLLER VGEN_Vx(4:0) 8 DATA REGISTERS PORT B COUNTERS TIMERS PWM ACQUISITION CHAIN VMULT (ZoomingADC) VLD PB(7:0) VMULT AC_R(3:0) AC_A(7:0) CMPD PB(7:6) UART PB(1:0) PA(3:0) evn PB(7:4) EVN HANDLING USRT PB(5:4) irq IRQ HANDLING SPI SPI(3:0) Figure 1-1. Block schematic of the XE8802 circuit. The UART (universal asynchronous receiver/transmitter) contains a full hardware implementation of the asynchronous serial link. © Semtech 2006 www.semtech.com 1-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver The counters/timers/PWM can take their clocks from internal or external sources (on Port A) and can generate interrupts or events. The PWM is output on Port B. The VLD (voltage level detector) detects the battery end of life with respect to a programmable threshold. The CMPD contains a 4-channel comparator. It is intended to monitor analog or digital signals with very low power consumption. 1.2 Pin map The 02 can be delivered in different packages. The pin maps for the different packages are given below. PA(2) PD1(2) PA(3) PD1(3) PA(4) PD1(4) 30 PA(5) PD1(5) PA(6) PD1(6) PA(7) PD1(7) PB(0) 40 LCD_VR2 PB(1) PB(2) PB(3) PB(4) PB(5) PB(6) PB(7) SPI(0) SPI(1) SPI(2) SPI(3) 50 VPP 1.2.1 LQFP-100 VMULT LCD_IO(0) VREG LCD_IO(1) VBAT LCD_IO(2) XOUT VSS LCD_IO(3) 20 LCD_VR1 XIN LCD_IO(4) AC_R(2) LCD_IO(5) AC_R(3) AC_R(0) LCD_IO(6) LCD_IO(7) AC_R(1) 60 LCD_IO(8) AC_A(7) LCD_IO(9) AC_A(6) LCD_IO(10) AC_A(5) LCD_IO(11) AC_A(4) AC_A(3) LCD_IO(12) LCD_IO(13) 10 AC_A(2) LCD_IO(14) AC_A(1) LCD_IO(15) AC_A(0) TEST LCD_IO(16) LCD_IO(17) NRESET 70 LCD_IO(18) PD2(7) LCD_IO(19) PD2(6) LCD_IO(20) PD2(5) PD2(4) LCD_IO(21) PD2(1) PA(1) PD2(0) PD1(1) PA(0) PD1(0) NFASTREAD VSS VGEN_V3 VGEN_VB PD2(3) PD2(2) 1 90 VGEN_V2 VGEN_V1 VGEN_VA LCD_COM(0) LCD_COM(1) LCD_IO(31) LCD_IO(30) LCD_IO(29) LCD_IO(28) LCD_IO(27) LCD_IO(26) LCD_IO(25) LCD_IO(24) LCD_IO(23) 80 LCD_IO(22) Figure 1-2. LQFP-100 pin map Package pin 1 2 name PD2(3) PD2(4) Package pin 51 52 name LCD_VR2 LCD_IO(0) © Semtech 2006 www.semtech.com 1-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 PD2(5) PD2(6) PD2(7) NRESET TEST AC_A(0) AC_A(1) AC_A(2) AC_A(3) AC_A(4) AC_A(5) AC_A(6) AC_A(7) AC_R(1) AC_R(0) AC_R(3) AC_R(2) XIN VSS XOUT VBAT VREG VMULT PA(2) PD1(2) PA(3) PD1(3) PA(4) PD1(4) PA(5) PD1(5) PA(6) PD1(6) PA(7) PD1(7) PB(0) PB(1) PB(2) PB(3) PB(4) PB(5) PB(6) PB(7) SPI(0) SPI(1) SPI(2) SPI(3) VPP 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 LCD_IO(1) LCD_IO(2) LCD_IO(3) LCD_VR1 LCD_IO(4) LCD_IO(5) LCD_IO(6) LCD_IO(7) LCD_IO(8) LCD_IO(9) LCD_IO(10) LCD_IO(11) LCD_IO(12) LCD_IO(13) LCD_IO(14) LCD_IO(15) LCD_IO(16) LCD_IO(17) LCD_IO(18) LCD_IO(19) LCD_IO(20) LCD_IO(21) LCD_IO(22) LCD_IO(23) LCD_IO(24) LCD_IO(25) LCD_IO(26) LCD_IO(27) LCD_IO(28) LCD_IO(29) LCD_IO(30) LCD_IO(31) LCD_COM(1) LCD_COM(0) VGEN_VA VGEN_V1 VGEN_V2 VGEN_V3 VGEN_VB VSS NFASTREAD PD1(0) PA(0) PD1(1) PA(1) PD2(0) PD2(1) PD2(2) Table 1-1. Bonding plan of the LQFP-100 package (LQFP 100L 14x14mm thick 1.6 mm) © Semtech 2006 www.semtech.com 1-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 1.2.2 LQFP-80 Package pin 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 name NRESET TEST AC_A(0) AC_A(1) AC_A(2) AC_A(3) AC_A(4) AC_A(5) AC_A(6) AC_A(7) AC_R(1) AC_R(0) AC_R(3) AC_R(2) XIN VSS XOUT VBAT VREG VMULT PA(2)/ PD1(2) PA(3)/ PD1(3) PA(4)/ PD1(4) PA(5)/ PD1(5) PA(6)/ PD1(6) PA(7) PD1(7) PB(0) PB(1) PB(2) PB(3) PB(4) PB(5) PB(6) PB(7) SPI(0) SPI(1) SPI(2) SPI(3) VPP Package pin 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 name LCD_VR1 LCD_IO(4) LCD_IO(5) LCD_IO(6) LCD_IO(7) LCD_IO(8) LCD_IO(9) LCD_IO(10) LCD_IO(11) LCD_IO(12) LCD_IO(13) LCD_IO(14) LCD_IO(15) LCD_IO(16) LCD_IO(17) LCD_IO(18) LCD_IO(19) LCD_IO(20) LCD_IO(21) LCD_IO(22) LCD_IO(23) LCD_IO(24) LCD_IO(25) LCD_IO(26) LCD_IO(27) LCD_IO(28) LCD_IO(29) LCD_IO(30) LCD_IO(31) LCD_COM(1) LCD_COM(0) VGEN_VA VGEN_V1 VGEN_V2 VGEN_V3 VGEN_VB VSS NFASTREAD PA(0)/PD1(0) PA(1)/ PD1(1) Table 1-2. Bonding plan of the LQFP-80 package (LQFP 80L 14x14mm thick 1.6 mm) © Semtech 2006 www.semtech.com 1-6 PD1(2)/PA(2) PD1(3)/PA(3) PD1(4)/PA(4) PD1(5)/PA(5) PD1(6)/PA(6) PA(7) PD1(7) PB(0) PB(1) PB(2) PB(3) PB(4) PB(5) PB(6) PB(7) SPI(0) SPI(1) SPI(2) SPI(3) 30 LCD_VR1 40 VPP XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 20 VMULT LCD_IO(4) VREG LCD_IO(5) VBAT LCD_IO(6) XOUT LCD_IO(7) VSS LCD_IO(8) XIN LCD_IO(9) AC_R(2) LCD_IO(10) AC_R(3) LCD_IO(11) AC_R(0) LCD_IO(12) AC_R(1) 50 LCD_IO(13) 10 AC_A(7) LCD_IO(14) AC_A(6) LCD_IO(15) AC_A(5) LCD_IO(16) AC_A(4) LCD_IO(17) AC_A(3) LCD_IO(18) AC_A(2) LCD_IO(19) AC_A(1) LCD_IO(20) AC_A(0) TEST LCD_IO(21) PD1(1)/PA(1) PD1(0)/PA(0) NFASTREAD VSS VGEN_VB VGEN_V3 VGEN_V2 VGEN_V1 VGEN_VA LCD_COM(0) LCD_IO(31) LCD_COM(1) LCD_IO(30) LCD_IO(29) LCD_IO(28) LCD_IO(27) LCD_IO(26) LCD_IO(25) LCD_IO(24) NRESET 80 1 70 60 LCD_IO(23) LCD_IO(22) Figure 1-3. LQFP- 80 pin map © Semtech 2006 www.semtech.com 1-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver (2600, 5315) (2716, 5315) (2882, 5315) (3050, 5315) (3217, 5315) (3385, 5315) (3553, 5315) (3720, 5315) (3888, 5315) (4055, 5315) (4223, 5315) LCD_COM(1) VSS LCD_IO(31) LCD_IO(30) LCD_IO(29) LCD_IO(28) LCD_IO(27) LCD_IO(26) LCD_IO(25) LCD_IO(24) LCD_IO(23) LCD_COM(0) (2432, 5315) PD2(2) PD2(1) PD2(0) VSS PA(1) PD1(1) PA(0) PD1(0) VBAT NFASTREAD VSS VGEN_VB VGEN_V3 VGEN_V2 VGEN_V1 VGEN_VA ( 450, 5315) ( 565, 5315) ( 680, 5315) ( 795, 5315) ( 910, 5315) (1025, 5315) (1140, 5315) (1255, 5315) (1370, 5315) (1485, 5315) (1600, 5315) (1761, 5315) (1876, 5315) (1991, 5315) (2106, 5315) (2221, 5315) 1.3 Bare die VGEN_V3 (4634, 4911) LCD_IO(22) (4634, 4795) PD2(3) PD2(4) PD2(5) PD2(6) PD2(7) VSS NRESET TEST LCD_IO(21) (4634, 4628) LCD_IO(20) (4634, 4460) LCD_IO(19) (4634, 4292) LCD_IO(18) (4634, 4124) LCD_IO(17) (4634, 3957) LCD_IO(16) (4634, 3789) XEMICS (48, 4673) (48, 4558) (48, 4443) (48, 4328) (48, 4213) (48, 4098) (48, 3983) (48, 3868) LCD_IO(15) (4634, 3622) LCD_IO(14) (4634, 3454) VSS (4634, 3288) LCD_IO(13) (4634, 3171) AC_A(0) AC_A(1) AC_A(2) AC_A(3) AC_A(4) AC_A(5) AC_A(6) AC_A(7) VBAT AC_R(1) AC_R(0) AC_R(3) AC_R(2) XIN VSS XOUT VBAT VREG VMULT LCD_IO(12) (4634, 3004) LCD_IO(11) (4634, 2826) LCD_IO(10) (4634, 2659) LCD_IO(9) (4634, 2376) LCD_IO(8) (4634, 2209) LCD_IO(7) (4634, 2041) LCD_IO(6) (4634, 1873) LCD_IO(5) (4634, 1706) VBAT 5530µm PA(2) PD1(2) PA(3) PD1(3) PA(4) PD1(4) VSS PA(5) PD1(5) PA(6) PD1(6) PA(7) PD1(7) VBAT PB(0) PB(1) PB(2) PB(3) PB(4) PB(5) PB(6) VSS PB(7) SPI(0) SPI(1) SPI(2) SPI(3) VPP (4231, 48) 4580µm ( 450, 48) ( 588, 48) ( 726, 48) ( 864, 48) (1002, 48) (1140, 48) (1278,48) (1416, 48) (1554, 48) (1692, 48) (1830, 48) (1968, 48) (2106, 48) (2249, 48) (2382, 48) (2520, 48) (2658, 48) (2796, 48) (2934, 48) (3049, 48) (3164, 48) (3279, 48) (3394, 48) (3512, 48) (3627, 48) (3742, 48) (3857, 48) (48, 3016) (48, 2874) (48, 2732) (48, 2590) (48, 2448) (48, 2306) (48, 2164) (48, 2022) (48, 1880) (48, 1738) (48, 1596) (48, 1454) (48, 1312) (48, 1170) (48, 1026) (48, 882) (48, 738) (48, 594) (48, 450) LCD_IO(4) (4634, 1538) LCD_VR1 LCD_IO(3) (4634, 1372) (4634, 1249) LCD_IO(2) (4634, 966) LCD_IO(1) (4634, 799) LCD_IO(0) (4634, 631) LCD_VR2 (4634, 465) Figure 1-4. Bare die dimensions and pin locations. © Semtech 2006 www.semtech.com 1-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 1.4 Pin assignment The table below gives a short description of the different pin assignments. Pin Assignment VBAT VSS VREG VPP NFASTREAD NRESET TEST XIN/XOUT PA(7:0) PB(7:0) PD1(7:0) PD2(7:0) SPI(3:0) LCD_IO(29:0) LCD_IO(31:30) LCD_COM(1:0) LCD_VR1/LCD_VR2 VGEN_Vx AC_A(7:0) AC_R(3:0) VMULT Positive power supply Negative power supply Connection for the mandatory external capacitor of the voltage regulator High voltage supply for flash memory programming (NC in ROM versions) Selects the maximal operation speed and the flash memory size (NC in ROM versions) Resets the circuit when the voltage is low Sets the pin to flash programming mode Quartz crystal connections, also used for flash memory programming Parallel input port A pins Parallel I/O port B pins Parallel I/O port D1 pins Parallel I/O port D2 pins Serial SPI port or general purpose I/O port pins LCD segment driver or general purpose I/O port pins LCD segment driver or LCD back plane driver or general purpose I/O port pins LCD back plane driver pins LCD supply voltage LCD driver voltage generation pins Acquisition chain input pins Acquisition chain reference pins Connection for external capacitor of the voltage multiplier Table 1-3. Pin assignment Table 1-4 gives a more detailed pin map for the different pins in the different packages. It also indicates the possible I/O configuration of these pins. The indications in blue bold are the configuration at start-up. Please note that in the LQFP-80 package several functions are routed to the same package pins. These pins are indicated in red italic. PD2(3) PD2(4) PD2(5) PD2(6) PD2(7) NRESET TEST AC_A(0) AC_A(1) AC_A(2) AC_A(3) AC_A(4) AC_A(5) AC_A(6) AC_A(7) AC_R(1) POWER LCD SNAP PD X X X X X PU X X X X X X X OD AO AI third first second I/O configuration DO 1 2 3 4 5 6 7 8 9 10 11 Function DI 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 lqfp-80 lqfp-100 Pin number X X X X X X X X X X X X X X X X X X X X X © Semtech 2006 www.semtech.com 1-9 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 12 13 14 15 16 17 18 19 20 21 21 22 22 23 23 24 24 25 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 AC_R(0) AC_R(3) AC_R(2) XIN VSS XOUT VBAT VREG VMULT PA(2) PD1(2) PA(3) PD1(3) PA(4) PD1(4) PA(5) PD1(5) PA(6) PD1(6) PA(7) PD1(7) PB(0) PB(1) PB(2) PB(3) PB(4) PB(5) PB(6) PB(7) SPI(0) SPI(1) SPI(2) SPI(3) VPP LCD_VR2 LCD_IO(0) LCD_IO(1) LCD_IO(2) LCD_IO(3) LCD_VR1 LCD_IO(4) LCD_IO(5) LCD_IO(6) LCD_IO(7) LCD_IO(8) LCD_IO(9) LCD_IO(10) LCD_IO(11) LCD_IO(12) LCD_IO(13) LCD_IO(14) X X X X X X X X X X X CNTC CNTD PWM0 PWM1 USRT_S0 USRT_S1 UART_Tx UART_Rx CMPD(0) CMPD(1) CMPD(2) CMPD(3) X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X © Semtech 2006 www.semtech.com 1-10 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 79 80 80 LCD_IO(15) LCD_IO(16) LCD_IO(17) LCD_IO(18) LCD_IO(19) LCD_IO(20) LCD_IO(21) LCD_IO(22) LCD_IO(23) LCD_IO(24) LCD_IO(25) LCD_IO(26) LCD_IO(27) LCD_IO(28) LCD_IO(29) LCD_IO(30) LCD_IO(31) LCD_COM(1) LCD_COM(0) VGEN_VA VGEN_V1 VGEN_V2 VGEN_V3 VGEN_VB VSS NFASTREAD PD1(0) PA(0) PD1(1) PA(1) PD2(0) PD2(1) PD2(2) X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X CNTA CNTB X X X X X X X X X X X X X X X X X X X X Pin map table legend: blue bold: configuration at start up AI: analog input AO: analog output DI: digital input DO: digital output OD: nMOS open drain output PU: pull-up resistor PD: pull-down resistor SNAP: snap-to-rail function (see peripheral description for detailed description) POWER: power supply Table 1-4. Pin description table © Semtech 2006 www.semtech.com 1-11 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 2 XE8802 Performance 2.1 Absolute maximum ratings 2-2 2.2 Operating range 2-2 2.3 Current consumption 2-3 2.4 2.4.1 2.4.2 2.4.3 Operating speed Flash circuit version ROM circuit version with regulator on ROM circuit version with regulator off 2-5 2-5 2-6 2-7 2.5 Simplified supply selection criteria 2-8 © Semtech 2006 www.semtech.com 2-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 2.1 Absolute maximum ratings Table 2-1. Absolute maximal ratings Voltage applied to VBAT with respect to VSS Voltage applied to VPP with respect to VSS Voltage applied to all pins except VPP and VBAT Storage temperature (ROM device or unprogrammed flash device) Storage temperature (programmed flash device) Min. Max. Note -0.3 VBAT-0.3 VSS-0.3 -55 6.0 12 VBAT+0.3 150 V V V °C -40 85 °C Stresses beyond the absolute maximal ratings may cause permanent damage to the device. Functional operation at the absolute maximal ratings is not implied. Exposure to conditions beyond the absolute maximal ratings may affect the reliability of the device. 2.2 Operating range Table 2-2. Operating range for the flash device Voltage applied to VBAT with respect to VSS Voltage applied to VBAT with respect to VSS during the flash programming Voltage applied to VPP with respect to VSS Voltage applied to all pins except VPP and VBAT Operating temperature range Capacitor on VREG (flash version) Capacitor on VMULT 1. 2. 3. Min. Max. Note 2.4 4.5 5.5 5.5 V V 1 VBAT VSS -40 0.8 1.0 11.5 VBAT 85 1.2 3.0 V V °C µF nF 2 3 During the programming of the device the temperature must be between 10°C and 40°C. The capacitor on VREG is mandatory. The capacitor on VMULT is optional. The capacitor has to be present if the multiplier is enabled. The multiplier has to be enabled if VBAT<3.0V. Table 2-3. Operating range for the ROM device Min. Voltage applied to VBAT with respect to VSS Acquisition chain off Acquisition chain on VREG bypassed VREG on VMULT on VMULT off Voltage applied to all pins except VPP and VBAT Operating temperature range Capacitor on VREG Capacitor on VMULT 1. 2. 3. Max. Note 1.2 5.5 V 1.5 2.4 3.0 5.5 5.5 5.5 V V V VSS -40 0.1 1.0 VBAT 125 1.2 3.0 V °C µF nF 2 1 3 The capacitor may be omitted when VREG is connected to VBAT. The voltage reference for the LCD drivers starts operating at 1.5 V. The capacitor on VMULT is optional. The capacitor has to be present if the multiplier is enabled. The multiplier has to be enabled if VBAT<3.0V. © Semtech 2006 www.semtech.com 2-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver All specifications in this document are valid for the complete operating range unless otherwise specified. Table 2-4. Operating range of the Flash memory Min. Retention time at 85°C Retention time at 55°C Number of programming cycles 1. 2. 2.3 10 100 10 Max. Note years years 1 1 2 Valid only if programmed using a programming tool that is qualified Circuits can be programmed more than 10 times but in that case, the retention time is no longer guaranteed. Current consumption The tables below give the current consumption for the circuit in different configurations. The figures are indicative only and may change as a function of the actual software implemented in the circuit. Table 2-5 gives the current consumption for the flash version of the circuit. The peripherals are disabled. The parallel ports are configured in input with pull up. Their pins are not connected externally. Table 2-5. Typical current consumption of the XE8802M version (8k instructions flash memory) Operation mode CPU RC Xtal Consumption High speed CPU 1 MIPS 1 MHz Off Low speed CPU .1 MIPS 100 kHz Off Low power CPU 32 kIPS Off 32 kHz HALT Off 32 kHz 200 µA 320 µA 410 µA 310 µA 21 µA 33 µA 42 µA 7.5 µA 11.0 µA 14.5 µA 1.9 µA 2.4V <>5.5V, 27°C HALT Ready 32kHz 2.3 µA 2.4V <>5.5V, 27°C HALT 1 MHz Off 35 µA 2.4V <>5.5V, 27°C 15 µA 2 µA 2.4V <>5.5V, 27°C 2.4V <>5.5V, 27°C Low power time keeping Fast wake-up time keeping Immediate wakeup time keeping VLD static current CMPD static current 1. 2. 3. 4. comments Note 2.4V<>5.5V, 27°C 1 2 3 4 1 2 3 1 2 3 2.4V <>5.5V, 27°C 2.4V <>5.5V, 27°C Software without data access 100% low power RAM access 100% RAM access typical software © Semtech 2006 www.semtech.com 2-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Table 2-6. Current consumption of the XE8802R version (8k instructions ROM memory) Operation mode CPU RC Xtal Consumption High speed CPU Max. Speed CPU Low speed CPU Low power CPU Low voltage CPU Low power time keeping 1 MIPS 4 MIPS .1 MIPS 32 kIPS 32 kIPS HALT 1 MHz 4 MHz 100 kHz Off Off Off Off Off Off 32 kHz 32 kHz 32 kHz 200 800 21 7 1 1.3 comments Note 2.4V<>5.5V, 27°C 2.4V<>5.5V, 27°C 2.4V <>5.5V, 27°C 2.4V <>5.5V, 27°C 1.2V, 27°C 2.4V <>5.5V, 27°C 1 1 1 1 1 1. Software using MOVE instruction using internal CPU registers and peripheral registers Hints for low power operation: 1. Use the low power RAM instead of the RAM for all parameters that are accessed frequently. The average current consumption for the low power RAM is about 40 times lower than for the RAM. 2. Rather than using the circuit at low speed, it is better to use the circuit at higher speed and switch off the blocks when not needed. 3. The power consumption of the program memory is an important part of the overall power consumption. In case you intend to use a ROM version and power consumption is too high, please ask us to provide you with a circuit version with smaller ROM size. © Semtech 2006 www.semtech.com 2-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 2.4 Operating speed 2.4.1 Flash circuit version The speed of the flash devices is not highly dependent upon the supply voltage. However, by limiting the temperature range, the speed can be increased. The minimal guaranteed speed as a function of the supply voltage and maximal temperature operating temperature is given in Figure2-2. VBAT 2.4 - 5.5 V VREG 1uF VSS Figure 2-1. Supply configuration for flash circuit operation. NFASTREAD=1 85°C NFASTREAD=0 85°C NFASTREAD=1 45°C NFASTREAD=0 45°C 7 speed (MIPS) 6 5 4 3 2 1 0 2 3 4 supply voltage VBAT (V) 5 Figure 2-2. Guaranteed speed as a function of the supply voltage and maximal temperature. The maximal speed of the device depends on the setting of the NFASTREAD pin. If NFASTREAD=1, the full 8k instructions of the memory are available and the speed is limited to 2-3MIPS. If NFASTREAD=0, only 4k instructions are available but the speed is doubled. Important Note The circuit has to be programmed for the correct option of the NFASTREAD pin. If not, the execution of the software may not be correct. © Semtech 2006 www.semtech.com 2-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 2.4.2 ROM circuit version with regulator on For the ROM version, two possible operating modes exist: with and without voltage regulator. Using the voltage regulator, low power consumption will be obtained even with supply voltages above 2.4V. Without the voltage regulator (i.e. VREG short-circuited to VBAT), a higher speed can be obtained. VBAT 2.4 - 5.5 V VREG 100nF VSS Figure 2-3. Supply configuration for ROM circuit operation using the internal regulator. 85°C 45°C 125°C speed (MIPS) 8 6 4 2 0 2 2.5 3 3.5 4 4.5 5 5.5 supply voltage VBAT (V) Figure 2-4. Guaranteed speed as a function of supply voltage and for different maximal temperatures using the voltage regulator. © Semtech 2006 www.semtech.com 2-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 2.4.3 ROM circuit version with regulator off VBAT VREG 1.2 – 3.3 V VSS Figure 2-5. Supply configuration for ROM circuit operation by-passing the internal regulator. 85°C 45°C 1.5 2 125°C speed (MIPS) 8 6 4 2 0 1 2.5 3 3.5 supply voltage VBAT (V) Figure 2-6. Guaranteed speed as a function of supply voltage and for two temperature ranges when VREG=VBAT. Important Note Note that the acquisition chain will not operate if VBAT is below 2.4V. The internal reference voltage for the LCD will not operate below 1.5V. If the internal reference is not used, the LCD voltage generator and the LCD driver will operate down to 1.2V. The operation range of the different blocks is summarized in Figure 2-7. © Semtech 2006 www.semtech.com 2-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 2.5 • • • • • • • Simplified supply selection criteria MTP devices always require the capacitor on VREG and VREG cannot be shorted to VBAT on MTP devices. ROM devices can operate 1.5 V to 5.5 V with lowest current requirement with the capacitor on VREG, and VREG not shorted to VBAT. ROM devices can operate 1.2 V to 3.3 V at highest speed with VREG shorted to VBAT. If operation is always above 3.0 V, the capacitor on VMULT is not needed and VMULT can be always off. If the acquisition chain is used between 2.4 V and 3.0 V, then the capacitor on VMULT must be present and VMULT must be set on during operation below 3.0 V. The acquisition chain does not operate below 2.4 V. The internal reference voltage for the LCD does not operate below 1.5 V. acquisition chain VMULT off acquisition chain VMULT on MTP Vgen internal reference ROM (VBAT≠VREG) CPU parallel and serial ports LCD driver and Vgen (no int. ref.) RC and crystal oscillator VLD Comparators Counters and PWM ROM (VBAT=VREG) 0 1.2 1.5 2.4 3.0 5.5 VBAT (V) Figure 2-7. Operating voltage range of the different circuit blocks. MTP devices do not operate below 2.4V. ROM devices can operate in different voltage ranges if VREG and VBAT are short circuited or not. © Semtech 2006 www.semtech.com 2-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 3. CPU 3.1 3.2 3.3 CPU description CPU internal registers CPU instruction short reference 3-2 3-2 3-4 © Semtech 2006 www.semtech.com 3-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 3.1 CPU description The CPU of the XE8000 series is a low power RISC core. It has 16 internal registers for efficient implementation of the C compiler. Its instruction set is made up of 35 generic instructions, all coded on 22 bits, with 8 addressing modes. All instructions are executed in one clock cycle, including conditional jumps and 8x8 multiplication. The circuit therefore runs on 1 MIPS on a 1MHz clock. The CPU hardware and software description is given in the document “Coolrisc816 Hardware and Software Reference Manual”. A short summary is given in the following paragraphs. The good code efficiency of the CPU core makes it possible to compute a polynomial like Z = ( A0 + A1 ⋅ Y ) ⋅ X + B0 + B1 ⋅ Y in less than 300 clock cycles (software code generated by the XEMICS Ccompiler, all numbers are signed integers on 16 bits). 3.2 CPU internal registers As shown in Figure 3-1, the CPU has 16 internal 8-bit registers. Some of these registers can be concatenated to a 16-bit word for use in some instructions. The function of these registers is defined in Table 3-1. The status register stat (Table 3-2) is used to manage the different interrupt and event levels. An interrupt or an event can both be used to wake up after a HALT instruction. The difference is that an interrupt jumps to a special interrupt function whereas an event continues the software execution with the instruction following the HALT instruction. The program counter (PC) is a 16 bit register that indicates the address of the instruction that has to be executed. The stack (STn) is used to memorise the return address when executing subroutines or interrupt routines. ST4 ST3 ST1 ST2 r0 r1 Data memory r2 data bus 22bit CPU CPU internal registers Instruction memory instruction bus PC program counter stack r3 i0h i0l i1h i1l i2h i2l i3h i3l iph ipl stat a Figure 3-1. CPU internal registers © Semtech 2006 www.semtech.com 3-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Register name r0 r1 r2 r3 i0h i0l i1h i1l i2h i2l i3h i3l iph ipl stat a Register function general purpose general purpose general purpose data memory offset MSB of the data memory index i0 LBS of the data memory index i0 MSB of the data memory index i1 LBS of the data memory index i1 MSB of the data memory index i2 LBS of the data memory index i2 MSB of the data memory index i3 LBS of the data memory index i3 MSB of the program memory index ip LBS of the program memory index ip status register accumulator Table 3-1. CPU internal register definition bit 7 6 5 4 name IE2 IE1 GIE IN2 3 IN1 2 IN0 1 EV1 0 EV0 function enables (when 1) the interrupt request of level 2 enables (when 1) the interrupt request of level 1 enables (when 1) all interrupt request levels interrupt request of level 2. The interrupts labelled “low” in the interrupt handler are routed to this interrupt level. This bit has to be cleared when the interrupt is served. interrupt request of level 1. The interrupts labelled “mid” in the interrupt handler are routed to this interrupt level. This bit has to be cleared when the interrupt is served. interrupt request of level 0. The interrupts labelled “hig” in the interrupt handler are routed to this interrupt level. This bit has to be cleared when the interrupt is served. event request of level 1. The events labelled “low” in the event handler are routed to this event level. This bit has to be cleared when the event is served. event request of level 1. The events labelled “hig” in the event handler are routed to this event level. This bit has to be cleared when the event is served. Table 3-2. Status register description The CPU also has a number of flags that can be used for conditional jumps. These flags are defined in Table 3-3. symbol Z C V name zero carry overflow function Z=1 when the accumulator a content is zero This flag is used in shift or arithmetic operations. For a shift operation, it has the value of the bit that was shifted out (LSB for shift right, MSB for shift left). For an arithmetic operation with unsigned numbers: it is 1 at occurrence of an overflow during an addition (or equivalent). it is 0 at occurrence of an underflow during a subtraction (or equivalent). This flag is used in shift or arithmetic operations. For arithmetic or shift operations with signed numbers, it is 1 if an overflow or underflow occurs. Table 3-3. Flag description © Semtech 2006 www.semtech.com 3-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 3.3 CPU instruction short reference Table 3-4 shows a short description of the different instructions available on the Coolrisc816. The notation cc in the conditional jump instruction refers to the condition description as given in Table 3-6. The notation reg, reg1, reg2, reg3 refers to one of the CPU internal registers of Table 3-1. The notation eaddr and DM(eaddr) refer to one of the extended address modes as defined in Table 3-5. The notation DM(xxx) refers to the data memory location with address xxx. Instruction Modification Operation Jump addr[15:0] Jump ip Jcc addr[15:0] Jcc ip Call addr[15:0] Call ip Calls addr[15:0] Calls ip -,-,-, -,-,-, -,-,-, -,-,-, -,-,-, -,-,-, -,-,-, -,-,-, - PC := addr[15:0] PC := ip if cc is true then PC := addr[15:0] if cc is true then PC := ip STn+1 := STn (n>1); ST1 := PC+1; PC := addr[15:0] STn+1 := STn (n>1); ST1 := PC+1; PC := ip ip := PC+1; PC := addr[15:0] ip := PC+1; PC := ip Ret Rets Reti Push Pop -,-,-, -,-,-, -,-,-, -,-,-, -,-,-, - PC := ST1; STn := STn+1 (n>1) PC := ip PC := ST1; STn := STn+1 (n>1); GIE :=1 PC := PC+1; STn+1 := STn (n>1); ST1 := ip PC := PC+1; ip := ST1; STn := STn+1 (n>1) Move reg,#data[7:0] Move reg1, reg2 Move reg, eaddr Move eaddr, reg Move addr[7:0],#data[7:0] -,-, Z, a -,-, Z, a -,-, Z, a -,-,-, -,-,-, - a := data[7:0]; reg := data[7:0] a := reg2; reg1 := reg2 a := DM(eaddr); reg := DM(eaddr) DM(eaddr) := reg DM(addr[7:0]) := data[7:0] Cmvd reg1, reg2 Cmvd reg, eaddr Cmvs reg1, reg2 Cmvs reg, eaddr -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a a := reg2; if C=0 then reg1 := a; a := DM(eaddr); if C=0 then reg := a a := reg2; if C=1 then reg1 := a; a := DM(eaddr); if C=1 then reg := a Shl reg1, reg2 Shl reg Shl reg, eaddr Shlc reg1, reg2 Shlc reg Shlc reg, eaddr Shr reg1, reg2 Shr reg Shr reg, eaddr Shrc reg1, reg2 Shrc reg Shrc reg, eaddr Shra reg1, reg2 Shra reg Shra reg, eaddr C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a a := reg2<<1; a[0] := 0; C := reg2[7]; reg1 := a a := reg<<1; a[0] := 0; C := reg[7]; reg := a a := DM(eaddr)<<1; a[0] :=0; C := DM(eaddr)[7]; reg := a a := reg2<<1; a[0] := C; C := reg2[7]; reg1 := a a := reg<<1; a[0] := C; C := reg[7]; reg := a a := DM(eaddr)<<1; a[0] := C; C := DM(eaddr)[7]; reg := a a := reg2>>1; a[7] := 0; C := reg2[0]; reg1 :=a a := reg>>1; a[7] := 0; C := reg[0]; reg := a a := DM(eaddr)>>1; a[7] := 0; C := DM(eaddr)[0]; reg := a a := reg2>>1; a[7] := C; C := reg2[0]; reg1 := a a := reg>>1; a[7] := C; C := reg[0]; reg := a a := DM(eaddr)>>1; a[7] := C; C := DM(eaddr)[0]; reg := a a := reg2>>1; a[7] := reg2[7]; C := reg2[0]; reg1 := a a := reg>>1; a[7] := reg[7]; C := reg[0]; reg := a a := DM(eaddr)>>1; a[7] := DM(eaddr)[7]; C := DM(eaddr)[0]; reg := a Cpl1 reg1, reg2 Cpl1 reg Cpl1 reg, eaddr Cpl2 reg1, reg2 Cpl2 reg Cpl2 reg, eaddr Cpl2c reg1, reg2 Cpl2c reg Cpl2c reg, eaddr -,-, Z, a -,-, Z, a -,-, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a a := NOT(reg2); reg1 := a a := NOT(reg); reg := a a := NOT(DM(eaddr)); reg := a a := NOT(reg2)+1; if a=0 then C:=1 else C := 0; reg1 := a a := NOT(reg)+1; if a=0 then C:=1 else C := 0; reg := a a := NOT(DM(eaddr))+1; if a=0 then C:=1 else C := 0; reg := a a := NOT(reg2)+C; if a=0 and C=1 then C:=1 else C := 0; reg1 := a a := NOT(reg)+C; if a=0 and C=1 then C:=1 else C := 0; reg := a a := NOT(DM(eaddr))+C; if a=0 and C=1 then C:=1 else C := 0; reg := a Inc reg1, reg2 Inc reg Inc reg, eaddr Incc reg1, reg2 Incc reg Incc reg, eaddr Dec reg1, reg2 C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a a := reg2+1; if a=0 then C := 1 else C := 0; reg1 := a a := reg+1; if a=0 then C := 1 else C := 0; reg := a a := DM(eaadr)+1; if a=0 then C := 1 else C := 0; reg := a a := reg2+C; if a=0 and C=1 then C := 1 else C := 0; reg1 := a a := reg+C; if a=0 and C=1 then C := 1 else C := 0; reg := a a := DM(eaadr)+C; if a=0 and C=1 then C := 1 else C := 0; reg := a a := reg2-1; if a=hFF then C := 0 else C := 1; reg1 := a © Semtech 2006 www.semtech.com 3-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Dec reg Dec reg, eaddr Decc reg1, reg2 Decc reg Decc reg, eaddr C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a a := reg-1; if a=hFF then C := 0 else C := 1; reg := a a := DM(eaddr)-1; if a=hFF then C := 0 else C := 1; reg := a a := reg2-(1-C); if a=hFF and C=0 then C := 0 else C := 1; reg1 := a a := reg-(1-C); if a=hFF and C=0 then C := 0 else C := 1; reg := a a := DM(eaddr)-(1-C); if a=hFF and C=0 then C := 0 else C := 1; reg := a And reg,#data[7:0] And reg1, reg2, reg3 And reg1, reg2 And reg, eaddr Or reg,#data[7:0] Or reg1, reg2, reg3 Or reg1, reg2 Or reg, eaddr Xor reg,#data[7:0] Xor reg1, reg2, reg3 Xor reg1, reg2 Xor reg, eaddr -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a -,-, Z, a a := reg and data[7:0]; reg := a a := reg2 and reg3; reg1 := a a := reg1 and reg2; reg1 := a a := reg and DM(eaddr); reg := a a := reg or data[7:0]; reg := a a := reg2 or reg3; reg1 := a a := reg1 or reg2; reg1 := a a := reg or DM(eaddr); reg := a a := reg xor data[7:0]; reg := a a := reg2 xor reg3; reg1 := a a := reg1 xor reg2; reg1 := a a := reg or DM(eaddr); reg := a Add reg,#data[7:0] Add reg1, reg2, reg3 Add reg1, reg2 Add reg, eaddr Addc reg,#data[7:0] Addc reg1, reg2, reg3 Addc reg1, reg2 Addc reg, eaddr Subd reg,#data[7:0] Subd reg1, reg2, reg3 Subd reg1, reg2 Subd reg, eaddr Subdc reg,#data[7:0] Subdc reg1, reg2, reg3 Subdc reg1, reg2 Subdc reg, eaddr Subs reg,#data[7:0] Subs reg1, reg2, reg3 Subs reg1, reg2 Subs reg, eaddr Subsc reg,#data[7:0] Subsc reg1, reg2, reg3 Subsc reg1, reg2 Subsc reg, eaddr C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a a := reg+data[7:0]; if overflow then C:=1 else C := 0; reg := a a := reg2+reg3; if overflow then C:=1 else C := 0; reg1 := a a := reg1+reg2; if overflow then C:=1 else C := 0; reg1 := a a := reg+DM(eaddr); if overflow then C:=1 else C := 0; reg := a a := reg+data[7:0]+C; if overflow then C:=1 else C := 0; reg := a a := reg2+reg3+C; if overflow then C:=1 else C := 0; reg1 := a a := reg1+reg2+C; if overflow then C:=1 else C := 0; reg1 := a a := reg+DM(eaddr)+C; if overflow then C:=1 else C := 0; reg := a a := data[7:0]-reg; if underflow then C := 0 else C := 1; reg := a a := reg2-reg3; if underflow then C := 0 else C := 1; reg1 := a a := reg2-reg1; if underflow then C := 0 else C := 1; reg1 := a a := DM(eaddr)-reg; if underflow then C := 0 else C := 1; reg := a a := data[7:0]-reg-(1-C); if underflow then C := 0 else C := 1; reg := a a := reg2-reg3-(1-C); if underflow then C := 0 else C := 1; reg1 := a a := reg2-reg1-(1-C); if underflow then C := 0 else C := 1; reg1 := a a := DM(eaddr)-reg-(1-C); if underflow then C := 0 else C := 1; reg := a a := reg-data[7:0]; if underflow then C := 0 else C := 1; reg := a a := reg3-reg2; if underflow then C := 0 else C := 1; reg1 := a a := reg1-reg2; if underflow then C := 0 else C := 1; reg1 := a a := reg-DM(eaddr); if underflow then C := 0 else C := 1; reg := a a := reg-data[7:0]-(1-C); if underflow then C := 0 else C := 1; reg := a a := reg3-reg2-(1-C); if underflow then C := 0 else C := 1; reg1 := a a := reg1-reg2-(1-C); if underflow then C := 0 else C := 1; reg1 := a a := reg-DM(eaddr)-(1-C); if underflow then C := 0 else C := 1; reg := a Mul reg,#data[7:0] Mul reg1, reg2, reg3 Mul reg1, reg2 Mul reg, eaddr Mula reg,#data[7:0] Mula reg1, reg2, reg3 Mula reg1, reg2 Mula reg, eaddr u, u, u, a u, u, u, a u, u, u, a u, u, u, a u, u, u, a u, u, u, a u, u, u, a u, u, u, a a := (data[7:0]*reg)[7:0]; reg := (data[7:0]*reg)[15:8] a := (reg2*reg3)[7:0]; reg1 := (reg2*reg3)[15:8] a := (reg2*reg1)[7:0]; reg1 := (reg2*reg1)[15:8] a := (DM(eaddr)*reg)[7:0]; reg := (DM(eaddr)*reg)[15:8] a := (data[7:0]*reg)[7:0]; reg := (data[7:0]*reg)[15:8] a := (reg2*reg3)[7:0]; reg1 := (reg2*reg3)[15:8] a := (reg2*reg1)[7:0]; reg1 := (reg2*reg1)[15:8] a := (DM(eaddr)*reg)[7:0]; reg := (DM(eaddr)*reg)[15:8] Mshl reg,#shift[2:0] Mshr reg,#shift[2:0] Mshra reg,#shift[2:0] u, u, u, a u, u, u, a u, u, u, a* a := (reg*2 )[7:0]; reg := (reg*2 )[15:8] (8-shift (8-shift a := (reg*2 )[7:0]; reg := (reg*2 )[15:8] (8-shift (8-shift a := (reg*2 )[7:0]; reg := (reg*2 )[15:8] Cmp reg,#data[7:0] Cmp reg1, reg2 Cmp reg, eaddr Cmpa reg,#data[7:0] Cmpa reg1, reg2 Cmpa reg, eaddr C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a C, V, Z, a a := data[7:0]-reg; if underflow then C :=0 else C:=1; V := C and (not Z) a := reg2-reg1; if underflow then C :=0 else C:=1; V := C and (not Z) a := DM(eaddr)-reg; if underflow then C :=0 else C:=1; V := C and (not Z) a := data[7:0]-reg; if underflow then C :=0 else C:=1; V := C and (not Z) a := reg2-reg1; if underflow then C :=0 else C:=1; V := C and (not Z) a := DM(eaddr)-reg; if underflow then C :=0 else C:=1; V := C and (not Z) Tstb reg,#bit[2:0] Setb reg,#bit[2:0] Clrb reg,#bit[2:0] Invb reg,#bit[2:0] -, -, Z, a -, -, Z, a -, -, Z, a -, -, Z, a a[bit] := reg[bit]; other bits in a are 0 reg[bit] := 1; other bits unchanged; a := reg reg[bit] := 0; other bits unchanged; a := reg reg[bit] := not reg[bit]; other bits unchanged; a := reg shift shift © Semtech 2006 www.semtech.com 3-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Sflag -,-,-, a a[7] := C; a[6] := C xor V; a[5] := ST full; a[4] := ST empty Rflag reg Rflag eaddr C, V, Z, a C, V, Z, a a := reg << 1; ; a[0] := 0; C := reg[7] a := DM(eaddr)<<1; a[0] :=0; C := DM(eaddr)[7] Freq divn -,-,-, - reduces the CPU frequency (divn=nodiv, div2, div4, div8, div16) Halt -,-,-, - halts the CPU Nop -,-,-, - no operation - = unchanged, u = undefined, *MSHR reg,# 1 doesn’t shift by 1 Table 3-4. Instruction short reference The Coolrisc816 has 8 different addressing modes. These modes are described in Table 3-5. In this table, the notation ix refers to one of the data memory index registers i0, i1, i2 or i3. Using eaddr in an instruction of Table 3-4 will access the data memory at the address DM(eaddr) and will simultaneously execute the index operation. extended address eaddr addr[7:0] (ix) (ix, offset[7:0]) (ix,r3) (ix)+ (ix,offset[7:0])+ -(ix) -(ix,offset[7:0]) accessed data memory location DM(eaddr) DM(h00&addr[7:0]) DM(ix) DM(ix+offset) DM(ix+r3) DM(ix) DM(ix+offset) DM(ix-1) DM(ix-offset) index operation ix := ix+1 ix := ix+offset ix := ix-1 ix := ix -offset direct addressing indexed addressing indexed addressing with immediate offset indexed addressing with register offset indexed addressing with index post-increment indexed addressing with index post-increment by the offset indexed addressing with index pre-decrement indexed addressing with index pre-decrement by the offset Table 3-5. Extended address mode description Eleven different jump conditions are implemented as shown in Table 3-6. The contents of the column CC in this table should replace the CC notation in the instruction description of Table 3-4. CC condition CS CC ZS ZC VS VC EV C=1 C=0 Z=1 Z=0 V=1 V=0 (EV1 or EV0)=1 After CMP op1,op2 EQ NE GT GE LT LE op1=op2 op1≠op2 op1>op2 op1≥op2 op1<op2 op1≤op2 Table 3-6. Jump condition description © Semtech 2006 www.semtech.com 3-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4 Memory Mapping 4.1 Memory organisation 4-2 4.2 Instruction memory and NFASTREAD 4-2 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.3.5 4.3.6 4.3.7 4.3.8 4.3.9 4.3.10 4.3.11 4.3.12 4.3.13 4.3.14 4.3.15 4.3.16 4.3.17 4.3.18 4.3.19 4.3.20 Quick reference data memory register map Low power data registers (h0000-h0007) System, clock configuration and reset configuration (h0010-h001F) Port A (h0020-h0027) Port B (h0028-h002F) Port D1 (h0030-h0033) Port D2 (h0034-h0037) Flash programming (h0038-003B) Event handler (h003C-h003F) Interrupt handler (h0040-h0047) USRT (h0048-h004F) UART (h0050-h0057) Counter/Timer/PWM registers (h0058-h005F) Acquisition chain registers (h0060-h0067) SPI registers (h0068-h006F) LCD voltage generator registers (h0070) Comparator registers (h0072-h0073) Voltage multiplier (h007C) Voltage Level Detector registers (h007E-h007F) RAM (h0080-h047F) LCD driver (h8000-8022) © Semtech 2006 4-3 4-4 4-4 4-5 4-5 4-5 4-6 4-6 4-6 4-7 4-7 4-8 4-8 4-8 4-9 4-9 4-9 4-9 4-9 4-9 4-10 www.semtech.com 4-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4.1 Memory organisation The XE8802 CPU is built with Harvard architecture. Harvard architecture uses separate instruction and data memories. The instruction bus and data bus are also separated. The advantage of such a structure is that the CPU can get a new instruction and read/write data simultaneously. The circuit configuration is shown in Figure 4-1. The CPU has its 16 internal registers. The instruction memory has a capacity of 8192 22-bit instructions if the pin NFASTREAD=1. The instruction memory has a capacity of 4096 22-bit instructions if the pin NFASTREAD=0. The data memory space has 8 low power registers, the peripheral register space, 1024 bytes of RAM and the LCD control register space. µF0h1FFF µF 0h8022 LCD registers CPU 4k x 22bit NFASTREAD=0 capacity: 1024 bytes r1 r2 r3 i0h i0l i1h i1l i2h i2l i3h i3l 0h007F Peripheral registers iph ipl 0h0008 0h0080 stat Data memory NFASTREAD=1 RAM r0 data bus capacity: 8k x 22bit 0h047F CPU internal registers Instruction memory instruction bus 0h8000 Low power RAM 0h0000 a Figure 4-1. Memory mapping The CPU internal registers are described in the CPU chapter. A short reference of the low power registers and peripheral registers is given in 4.3. 4.2 Instruction memory and NFASTREAD As indicated in chapter 4.1, the instruction memory can be organized in two ways: 4096 instructions with a speed of 4 to 6 MIPS or 8192 instructions with a speed of 2 to 3 MIPS (see XE8802 performance chapter for more details). Important Note: When programming the device, the correct NFASTREAD option has to be selected in the programmer. Failure to do so may result in non-functional software. The default option in the programme is NFASTREAD=1, i.e. the 8k instruction slower operation mode. © Semtech 2006 www.semtech.com 4-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Important Note: In the actual version of the Prostart, the NFASTREAD pin is hard wired to VBAT. Nevertheless, the Prostart can be used to program the device for both modes by (un)checking the NFASTREAD selection box in the software loader. If the user wants to run the software programmed in the NFASTREAD=0 mode, he will need to build his own application board which connects NFASTREAD=VSS. 4.3 Quick reference data memory register map The data register map is given in the tables below. A more detailed description of the different registers is given in the detailed description of the different peripherals. The tables give the following information: 1. The register name and register address 2. The different bits in the register 3. The access mode of the different bits (see Table 4-4-1 for code description) 4. The reset source and reset value of the different bits The reset source coding is given in Table 4-4-2. To get a full description of the reset sources, please refer to the reset block chapter. code r w r0 r1 c c1 ca s access mode bit can be read bit can be written bit always reads 0 bit always reads 1 bit is cleared by writing any value bit is cleared by writing a 1 bit is cleared after reading special function, verify the detailed description in the respective peripherals Table 4-4-1. Access mode codes used in the register definitions code glob cold pconf sleep reset source nresetglobal nresetcold nresetpconf nresetsleep Table 4-4-2. Reset source coding used in the register definitions © Semtech 2006 www.semtech.com 4-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4.3.1 Low power data registers (h0000-h0007) Name Address 7 6 5 Reg00 h0000 Reg01 h0001 Reg02 h0002 Reg03 h0003 Reg04 h0004 Reg05 h0005 Reg06 h0006 Reg07 h0007 4 3 Reg00[7:0] rw, 00000000, glob Reg01[7:0] rw,00000000,glob Reg02[7:0] rw,00000000,glob Reg03[7:0] rw,00000000,glob Reg04[7:0] rw,00000000,glob Reg05[7:0] rw,00000000,glob Reg06[7:0] rw,00000000,glob Reg07[/:0] rw,0000000,glob 2 1 0 Table 4-4-3. Low power data registers 4.3.2 System, clock configuration and reset configuration (h0010-h001F) Name Address 7 RegSysCtrl SleepEn h0010 rw,0,cold RegSysReset Sleep h0011 rw,0,glob RegSysClock CpuSel h0012 rw,0,sleep RegSysMisc h0013 r0 RegSysWd h0014 r0 RegSysPre0 h0015 r0 RegSysRcTrim1 h001B RegSysRcTrim2 h001C 6 ResetfromportA r0 rc, 0, cold EnExtClock rw,0,cold 4 EnResetWD rw,0,cold ResetWD rc, 0, cold BiasRC rw,1,cold r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 RcFreqRange rw,0,cold EnResetPConf rw,0,cold SleepFlag rc,0,cold r0 r0 r0 r0 5 EnBusError rw,0,cold ResetBusError 3 2 1 0 r0 r0 r0 r0 rc, 0, cold ColdXtal r,1,sleep r0 r0 r0 r0 EnableXtal EnableRC r0 rw,0,sleep rw,1,sleep Output16k OutputCpuCk r0 rw,0,sleep rw,0,sleep WatchDog[3:0] s,0000,glob r0 r0 ClearLowPresca l c1r0,0,- RcFreqCoarse[3:0] rw,0001,cold RcFreqFine[5:0] rw,00000,cold Table 4-4-4. Reset block and clock block registers © Semtech 2006 www.semtech.com 4-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4.3.3 Port A (h0020-h0027) Name Address 7 6 5 3 PAIn[7:0] r PADebounce[7:0] rw,00000000,pconf PAEdge[7:0] rw,00000000,glob PAPullup[7:0] rw,11111111,pconf PARes0[7:0] rw, 00000000, glob PARes1[7:0] rw,00000000,glob 2 1 0 r0 r0 r0 r0 r0 PASnaptorail[7:0] rw,00000000,pconf r0 r0 DebFast rw,0,pconf RegPAIn h0020 RegPADebounce h0021 RegPAEdge h0022 RegPAPullup h0023 RegPARes0 h0024 RegPARes1 h0025 RegPACtrl h0026 RegPASnaptorail h0027 4 Table 4-4-5. Port A registers 4.3.4 Port B (h0028-h002F) Name Address 7 6 5 3 PBOut[7:0] rw,00000000,pconf PBIn[7:0] r PBDir[7:0] rw,00000000,pconf PBOpen[7:0] rw,00000000,pconf PBPullup[7:0] rw,11111111,pconf PBAna[7:0] rw,00000000,pconf 2 1 0 5 4 3 PD1Out[7:0] rw,00000000,pconf PD1In[7:0] r PD1Dir[7:0] rw,00000000,pconf 2 1 0 RegPBOut h0028 RegPBIn h0029 RegPBDir h002A RegPBOpen h002B RegPBPullup h002C RegPBAna h002D 4 Table 4-4-6. Port B registers 4.3.5 Port D1 (h0030-h0033) Name Address RegPD1Out h0030 RegPD1In h0031 RegPD1Dir h0032 RegPD1Pullup h0033 7 6 PD1SnapToRail[3:0] rw,0000,pconf PD1Pullup[3:0] rw,1111,pconf Table 4-4-7. Port D1 registers © Semtech 2006 www.semtech.com 4-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4.3.6 Port D2 (h0034-h0037) Name Address RegPD2Out h0034 RegPD2In h0035 RegPD2Dir h0036 RegPD2Pullup h0037 7 6 5 4 3 PD2Out[7:0] rw,00000000,pconf PD2In[7:0] r PD2Dir[7:0] rw,00000000,pconf PD2SnapToRail[3:0] rw,0000,pconf 2 1 0 PD2Pullup[3:0] rw,1111,pconf Table 4-4-8. Port D2 registers 4.3.7 Flash programming (h0038-003B) These four registers are used during flash programming only. Refer to the flash programming algorithm documentation for more details. 4.3.8 Event handler (h003C-h003F) Name Address RegEvn h003C RegEvnEn h003D RegEvnPriority h003E RegEvnEvn h003F 7 CntIrqA rc1,0,glob 6 CntIrqC rc1,0,glob 5 128Hz rc1,0,glob r0 r0 r0 4 3 PAEvn[1] CntIrqB rc1,0,glob rc1,0,glob EvnEn[7:0] rw,00000000,glob EvnPriority[7:0] r,11111111,glob r0 r0 2 CntIrqD rc1,0,glob 1 1Hz rc1,0,glob 0 PAEvn[0] rc1,0,glob r0 EvnHigh r,0,glob EvnLow r,0,glob Table 4-4-9. Event handler registers The origin of the different events is summarised in the table below. Event CntIrqA CntIrqB CntIrqC CntIrqD 128Hz 1Hz PAEvn[1:0] Event source Counter/Timer A (counter block) Counter/Timer B (counter block) Counter/Timer C (counter block) Counter/Timer D (counter block) Low prescaler (clock block) Low prescaler (clock block) Port A Table 4-4-10. Event source description © Semtech 2006 www.semtech.com 4-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4.3.9 Interrupt handler (h0040-h0047) Name Address 7 IrqAC h0040 rc1,0,glob RegIrqMid UsrtCond2 h0041 rc1,0,glob RegIrqLow PAIrq[7] h0042 rc1,0,glob RegIrqEnHig h0043 RegIrqEnMid h0044 RegIrqEnLow h0045 RegIrqPriority h0046 RegIrqIrq h0047 r0 RegIrqHig 6 128Hz rc1,0,glob UrstCond1 rc1,0,glob PAIrq[6] rc1,0,glob 5 IrqSPI rc1,0,glob PAIrq[5] rc1,0,glob CntIrqB rc1,0,glob r0 r0 4 3 CntIrqA CntIrqC rc1,0,glob rc1,0,glob PAIrq[4] 1Hz rc1,0,glob rc1,0,glob CntIrqD PAIrq[3] rc1,0,glob rc1,0,glob IrqEnHig[7:0] rw,0000000,glob IrqEnMid[7:0] rw,0000000,glob IrqEnLow[7:0] rw,0000000,glob IrqPriority[7:0] r,11111111,glob r0 r0 2 CmpdIrq rc1,0,glob VldIrq rc1,0,glob PAIrq[2] rc1,0,glob 1 UartIrqTx rc1,0,glob PAIrq[1] rc1,0,glob 0 UartIrqRx rc1,0,glob PAIrq[0] rc1,0,glob r0 r0 IrqHig r,0,glob IrqMid r,0,glob IrqLow r,0,glob 1 Table 4-4-11. Interrupt handler registers The origin of the different interrupts is summarised in the table below. Event CmpdIrq CntIrqA CntIrqB CntIrqC CntIrqD 128Hz 1Hz PAIrq[7:0] UartIrqRx UartIrqTx UrstCond1 UsrtCond2 VldIrq IrqAC IrqSPI Event source Low power comparators Counter/Timer A (counter block) Counter/Timer B (counter block) Counter/Timer C (counter block) Counter/Timer D (counter block) Low prescaler (clock block) Low prescaler (clock block) Port A UART reception UART transmission USRT condition 1 USRT condition 2 Voltage level detector Acquisition chain end of conversion interrupt SPI end of reception/transmission interrupt Table 4-4-12. Interrupt source description 4.3.10 USRT (h0048-h004F) Name Address 7 6 5 4 3 2 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 UsrtEnWaitCond1 RegUsrtS1 h0048 RegUsrtS0 h0049 RegUsrtCond1 h004A RegUsrtCond2 h004B RegUsrtCtrl h004C RegUsrtBufferS1 h004D RegUsrtEdgeS0 h004E r0 r0 r0 r0 r0 UsrtWaitS0 r,0,glob r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 r0 rw,0,glob 0 UsrtS1 r0 s,1,glob UsrtS0 r0 s,1,glob UsrtCond1 r0 rc,0,glob UsrtCond2 r0 rc,0,glob UsrtEnWaitS0 UsrtEnable rw,0,glob rw,0,glob UsrtBufferS1 r0 r,0,glob UsrtEdgeS0 r0 r,0,glob Table 4-4-13. USRT register description © Semtech 2006 www.semtech.com 4-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4.3.11 UART (h0050-h0057) Name Address RegUartCtrl h0050 RegUartCmd h0051 RegUartTx h0052 RegUartTxSta h0053 RegUartRx h0054 RegUartRxSta h0055 7 UartEcho rw,0,glob SelXtal rw,0,glob 6 UartEnRx rw,0,glob r0 r0 r0 r0 5 UartEnTx rw,0,glob r0 4 3 UartXRx UartXTx rw,0,glob rw,0,glob UartRcSel[2:0] rw,000,glob UartTx[7:0] rw,0000000,glob 2 UartPM rw,0,glob 1 UartBR[2:0] rw,101,glob UartPE rw,0,glob 0 UartWL rw,1,glob UartTxBusy UartTxFull r0 r0 r0 r,0,glob r,0,glob UartRx[7:0] r,00000000,glob UartRxSErr UartRxPErr UartRxFErr UartRxOerr UartRxBusy UartRxFull r,0,glob r,0,glob r,0,glob rc,0,glob r,0,glob r,0,glob r0 Table 4-14. UART register description 4.3.12 Counter/Timer/PWM registers (h0058-h005F) Name Address 7 RegCntA h0058 RegCntB h0059 RegCntC h005A RegCntD h005B RegCntCtrlCk h005C RegCntConfig1 h005D RegCntConfig2 h005E RegCntOn h005F 6 5 4 3 2 1 0 CounterA[7:0] s,00000000,glob CounterB[7:0] s,00000000,glob CounterC[7:0] s,00000000,glob CounterD[7:0] s,00000000,glob CntDCkSel[1:0] CntCCkSel[1:0] CntBCkSel[1:0] CntACkSel[1:0] rw,00,glob rw,00,glob rw,00,glob rw,00,glob CntDDownUp CntCDownUp CntBDownUp CntADownUp CascadeCD CascadeAB CntPWM1 CntPWM0 rw,0,glob rw,0,glob rw,0,glob rw,0,glob rw,0,glob rw,0,glob rw,0,glob rw,0,glob CapSel[1:0] CapFunc[1:0] Pwm1Size[1:0] Pwm0Size[1:0] rw,00,glob rw,00,glob rw,00,glob rw,00,glob CntDExtDiv CntCExtDiv CntBExtDiv CntAExtDiv CntDEnable CntCEnable CntBEnable CntAEnable rw,0,glob rw,0,glob rw,0,glob rw,0,glob rw,0,glob rw,0,glob rw,0,glob rw,0,glob Table 4-15. Counter/timer/PWM register description. 4.3.13 Acquisition chain registers (h0060-h0067) Name Address 7 6 5 4 3 2 1 RegAcOutLsb OUT[7:0] h0060 r,0,glob RegAcOutMsb OUT[15:8] h0061 r,0,glob RegAcCfg0 START SET_NELCONV[1:0] SET_OSR[2:0] CONT h0062 w r0,0,glob rw,01,glob rw,010,glob rw,0,glob RegAcCfg1 IB_AMP_ADC[1:0] IB_AMP_PGA[1:0] ENABLE[3:0] h0063 rw,11,glob rw,11,glob rw,0000,glob RegAcCfg2 FIN PGA2_GAIN[1:0] PGA2_OFFSET[3:0] h0064 rw,00,glob rw,00,glob rw,0000,glob RegAcCfg3 PGA1_GAIN PGA3_GAIN[6:0] h0065 rw,0000000,glob Rw,0,glob RegAcCfg4 PGA3_OFFSET h0066 r0 rw,0000000,glob RegAcCfg5 BUSY DEF AMUX[4:0] h0067 r,0,glob w r0 rw,00000,glob 0 r0 VMUX rw,0,glob Table 4-16. Acquisition chain register description. © Semtech 2006 www.semtech.com 4-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4.3.14 SPI registers (h0068-h006F) Name Address 7 6 RegSpiControl ClearCounter NotSlaveSelect rw,1,glob h0068 c1 r0 RegSpiStatus h0069 r0 r0 RegSpiDataOut h006A RegSpiDataIn h006B RegSpiPullup h006C r0 r0 RegSpiDir h006D r0 r0 5 SpiMaster rw,1,glob r0 r0 r0 4 3 2 1 0 SpiEnable ClockPhase ClockPolarity BaudRate[1:0] rw,0,glob rw,1,glob rw,00,glob rw,0,glob SpiOverflow SpiRxFull SpiTxEmpty r0 r0 r c1,0,glob r,0,glob r w1,1,glob SpiDataOut[7:0] rw,00000000,glob SpiDataIn[7:0] r,00000000,glob SpiPullup[3:0] r0 rw,1111,pconf SpiDir[3:0] r0 rw,0000,pconf Table 4-17. SPI register description. 4.3.15 LCD voltage generator registers (h0070) Name Address RegVgenCfg0 h0070 7 6 r0 r0 5 4 VgenClkSel[1:0] rw,10,glob 3 VgenOff rw,1,glob 2 VgenMode rw,0,glob 1 VgenStdb rw,0,glob 0 VgenRefEn rw,0,glob Table 4-18. LCD voltage generator register. 4.3.16 Comparator registers (h0072-h0073) Name Address RegCmpdStat h0072 RegCmpdCtrl h0073 7 6 5 CmpdStat[3:0] rca,0000,glob IrqOnRising[2:0] rw,000,glob 4 3 2 1 CmpdOut[3:0] r,0000,glob EnIrqCh[3:0] rw,0000,glob 0 Enable rw,0,glob Table 4-19. Low power comparator registers 4.3.17 Voltage multiplier (h007C) Name Address RegVmultCfg0 h007C 7 r0 6 r0 5 r0 4 3 r0 r0 2 Enable rw,0,glob 1 0 Fin[1:0] rw,00,glob Table 4-20. VMULT register. 4.3.18 Voltage Level Detector registers (h007E-h007F) Name Address 7 6 5 4 r0 r0 r0 r0 3 VldRange rw,0,glob r0 r0 r0 r0 r0 RegVldCtrl h007E RegVldStat h007F 2 1 VldTune[2:0] rw,000,glob VldResult VldValid r,0,glob r,0,glob 0 VldEn rw,0,glob Table 4-21. Voltage level detector register description 4.3.19 RAM (h0080-h047F) The 1024 RAM bytes can be accessed for read and write operations. The RAM has no reset function. Variables stored in the RAM should be initialised before use since they can have any value at circuit start up. © Semtech 2006 www.semtech.com 4-9 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4.3.20 LCD driver (h8000-8022) Name Address RegLcdData0 h8000 RegLcdData1 h8001 RegLcdData2 h8002 RegLcdData3 h8003 RegLcdData4 h8004 RegLcdData5 h8005 RegLcdData6 h8006 RegLcdData7 h8007 RegLcdData8 h8008 RegLcdData9 h8009 RegLcdData10 h800A RegLcdData11 h800B RegLcdData12 h800C RegLcdData13 h800D RegLcdData14 h800E RegLcdData15 h800F RegPLcdOut0 h8010 RegPLcdOut1 h8011 RegPLcdOut2 h8012 RegPLcdOut3 h8013 RegPLcdDir0 h8014 RegPLcdDir1 h8015 RegPLcdDir2 h8016 RegPLcdDir3 h8017 RegPLcdPullup0 h8018 RegPLcdPullup1 h8019 RegPLcdPullup2 h801A RegPLcdPullup3 h801B RegPLcdIn0 h801C RegPLcdIn1 h801D RegPLcdIn2 h801E RegPLcdIn3 h801F 7 6 5 4 3 LcdData0[7:0] rw,00000000,pconf LcdData1[7:0] rw,00000000,pconf LcdData2[7:0] rw,00000000,pconf LcdData3[7:0] rw,00000000,pconf LcdData4[7:0] rw,00000000,pconf LcdData5[7:0] rw,00000000,pconf LcdData6[7:0] rw,00000000,pconf LcdData7[7:0] rw,00000000,pconf LcdData8[7:0] rw,00000000,pconf LcdData9[7:0] rw,00000000,pconf LcdData10[7:0] rw,00000000,pconf LcdData11[7:0] rw,00000000,pconf LcdData12[7:0] rw,00000000,pconf LcdData13[7:0] rw,00000000,pconf LcdData14[7:0] rw,00000000,pconf LcdData15[7:0] rw,00000000,pconf PLcdOut0[7:0] rw,00000000,pconf PLcdOut1[7:0] rw,00000000,pconf PLcdOut2[7:0] rw,00000000,pconf PLcdOut3[7:0] rw,00000000,pconf PLcdDir0[7:0] rw,00000000,pconf PLcdDir1[7:0] rw,00000000,pconf PLcdDir2[7:0] rw,00000000,pconf PLcdDir3[7:0] rw,00000000,pconf PLcdPullup0[7:0] rw,00000000,pconf PLcdPullup1[7:0] rw,00000000,pconf PLcdPullup2[7:0] rw,00000000,pconf PLcdPullup3[7:0] rw,00000000,pconf PLcdIn0[7:0] r PLcdIn1[7:0] r PLcdIn2[7:0] r PLcdIn3[7:0] r © Semtech 2006 2 1 0 www.semtech.com 4-10 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver r0 LcdSe15 rw,1,glob r0 LcdSe19 rw,1,glob LcdSleep rw,1,glob LcdSe23 rw,1,glob r0 r0 r0 RegLcdOn h8020 RegLcdSe h8021 RegLcdClkFrame h8022 r0 LcdSe3 rw,1,glob r0 r0 LcdSe7 LcdSe11 rw,1,glob rw,1,glob LcdDivFreq[2:0] rw,000,glob LcdMux[1:0] rw,00,glob LcdSe27 LcdSe31 rw,1,glob rw,1,glob LcdFreq[1:0] rw,00,glob Table 4-22. LCD driver registers. © Semtech 2006 www.semtech.com 4-11 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 5. Low Power Modes 5.1 5.1.1 5.2 FEATURES ........................................................................................................................................ 5-2 Overview ......................................................................................................................................... 5-2 OPERATING MODE ............................................................................................................................. 5-2 © Semtech 2006 www.semtech.com 5-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 5.1 5.1.1 Features Overview The XE8000 chips have three operating modes. These are the normal, low current and very low current modes (see Figure 5-1). The different modes are controlled by the reset and clock blocks (see the documentation of the respective blocks). 5.2 Operating mode Start-up All bits are reset in the design when a POR or padnreset is active. RC is enabled, Xtal is disabled and CPU is reset (pmaddr = 0000). If the port A is used to return from the sleep mode, all bits with nresetcold do not change (see sleep mode) Start-up All bits with nresetglobal and nresetpconf(if enabled) are reset. Clock configuration doesn’t change except cpuck (freqdiv is reset, see clock block). CPU is reset Active mode This is the mode where the CPU and all peripherals can work and execute the embedded software. Standby mode Executing a HALT instruction moves the XE8000 into the Standby mode. The CPU is stopped, but the clocks remain active. Therefore, the enabled peripherals remain active e.g. for time keeping. A reset or an interrupt/event request (if enabled) cancels the standby mode. Sleep mode This is a very low-power mode because all circuit clocks and all peripherals are stopped. Only some service blocks remain active. No time-keeping is possible. Two instructions are necessary to move into sleep mode. First, the SleepEn (sleep enable) bit in RegSysCtrl has to be set to 1. The sleep mode can then be activated by setting the Sleep bit in RegSysReset to 1. There are three possibe ways to wake-up from the sleep mode: 1. The POR (power-on-reset caused by a power-down followed by power-on). The RAM information is lost. 2. The padnreset 3. The Port A reset combination (if the Port A is present in the product). See Port A documentation for more details. Note: If the Port A is used to return from the sleep mode, all bits with nresetcold do not change (RegSysCtrl, RegSysReset (except bit sleep), Enextclock and Biasrc in RegSysClock, RegSysRcTrim1 and RegSysRcTrim2). The SleepFlag bit in RegSysReset, reads back a 1 if the circuit was in sleep mode since the flag was last cleared (see reset block for more details). Note: It is recommended to insert a NOP instruction after the instruction that sets the circuit in sleep mode because this instruction can be executed when the sleep mode is left using the resetfromportA. © Semtech 2006 www.semtech.com 5-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver START-UP without condition por padnreset por padnreset RESET por padnreset without condition por padnreset portA reset portA reset watchdog reset portA reset watchdog reset buserror reset Halt instruction ACTIVE Interrupt/event STAND-BY SLEEP set bit sleep normal mode low current very low current Figure 5-1. XE8000 operating modes. © Semtech 2006 www.semtech.com 5-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 6. Reset Generator 6.1 6.2 6.3 6.4 6.5 6.5.1 6.5.2 6.5.3 6.5.4 6.5.5 6.6 6.7 6.8 6.9 FEATURES OVERVIEW REGISTER MAP RESET HANDLING CAPABILITIES RESET SOURCE DESCRIPTION Power On Reset NRESET pin Programmable Port A input combination Watchdog reset BusError reset SLEEP MODE CONTROL REGISTER DESCRIPTION AND OPERATION WATCHDOG START-UP AND WATCHDOG SPECIFICATIONS © Semtech 2006 6-2 6-2 6-2 6-3 6-4 6-4 6-4 6-4 6-4 6-5 6-5 6-5 6-5 6-6 www.semtech.com 6-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 6.1 • • • • • Features Power On Reset (POR) External reset from the NRESET pin Programmable Watchdog timer reset Programmable BusError reset Sleep mode management Product dependant: • Programmable Port A input combination reset 6.2 Overview The reset block is the reset manager. It handles the different reset sources and distributes them through the system. It also controls the sleep mode of the circuit. 6.3 Register map register name RegSysCtrl RegSysReset RegSysWd Table 6-1. Reset registers Table 6-1 gives the different registers used by this block. © Semtech 2006 www.semtech.com 6-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Pos. 7 RegSysCtrl SleepEn Rw rw Reset 0 nresetcold 6 EnResetPConf rw 0 nresetcold 5 EnBusError rw 0 nresetcold 4 EnResetWD rw 0 nresetcold - r 0000 3–0 Function enables Sleep mode 0: sleep mode is disabled 1: sleep mode is enabled enables the nresetpconf signal when the nresetglobal is active 0: nresetpconf is disabled 1: nresetpconf is enabled enables reset from BusError 0: BusError reset source is disabled 1: BusError reset source is enabled enables reset from Watchdog 0: Watchdog reset source is disabled 1: Watchdog reset source is enabled this bit can not be set to 0 by SW unused Table 6-2. RegSysCtrl register. Pos. 7 6 5 4 3 2–0 RegSysReset Sleep SleepFlag ResetBusError ResetWD ResetfromportA Rw rw rc rc rc rc r Reset 0 nresetglobal 0 nresetcold 0 nresetcold 0 nresetcold 0 nresetcold 000 Function Sleep mode control (reads always 0) Sleep mode was active before reset source was BusError reset source was Watchdog reset source was Port A combination unused Table 6-3. RegSysReset register Pos. 7-4 3 2 1 0 RegSysWd WDKey[3] WDCounter[3] WDKey[2] WDCounter[2] WDKey[1] WDCounter[1] WDKey[0] WDCounter[0] Rw r w r w r w r w r Reset 0000 Function unused 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal Watchdog Key bit 3 Watchdog counter bit 3 Watchdog Key bit 2 Watchdog counter bit 2 Watchdog Key bit 1 Watchdog counter bit 1 Watchdog Key bit 0 Watchdog counter bit 0 Table 6-4. RegSysWd register 6.4 Reset handling capabilities There are 5 reset sources: • Power On Reset (POR) • External reset from the NRESET pin • Programmable port A input combination • Programmable watchdog timer reset • Programmable BusError reset on processor access outside the allocated memory map © Semtech 2006 www.semtech.com 6-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Another reset source is the bit Sleep in the RegSysReset register. This source is fully controlled by software and is only used during the sleep mode. Four internal reset signals are generated from these sources and distributed through the system: • nresetcold: is asserted on POR or by the NRESET pin • nresetglobal: is asserted when nresetcold or any other enabled reset source is active • nresetsleep: is asserted when the circuit is in sleep mode • nresetpconf: is asserted when nresetglobal is active and if the EnResetPConf bit in the RegSysCtrl register is set. This reset is generally used in the different ports. It allows to maintain the port configuration unchanged while the rest of the circuit is reset. Table 6-5 shows a summary of the dependency of the internal reset signals on the various reset sources. In all the tables describing the different registers, the reset source is indicated. Internal reset signals Asserted reset source nresetglobal POR NRESET pin PortA input Watchdog BusError Sleep Asserted Asserted Asserted Asserted Asserted - nresetpconf when when EnResetPConf EnRestPConf is set to 0 is set to 1 Asserted Asserted Asserted Asserted Asserted Asserted Asserted - nresetsleep nresetcold Asserted Asserted Asserted Asserted Asserted - Table 6-5. Internal reset assertion as a function of the reset source. 6.5 6.5.1 Reset source description Power On Reset The power on reset (POR) monitors the external supply voltage. It activates a reset on a rising edge of this supply voltage. The reset is inactivated only if the internal voltage regulator has started up. The POR block performs no precise voltage level detection. 6.5.2 NRESET pin Applying a low input state on the NRESET pin can activate the reset. 6.5.3 Programmable Port A input combination Port A (if present in the product) can generate a reset signal. information. 6.5.4 See the description of the Port A for further Watchdog reset The Watchdog will generate a reset if the EnResetWD bit in the RegSysCtrl register has been set and if the watchdog is not cleared in time by the processor. See chapter 6.8 describing the watchdog for further information. © Semtech 2006 www.semtech.com 6-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 6.5.5 BusError reset The address space is assigned as shown in the register map of the product. If the EnBusError bit in the RegSysCtrl register is set and the software accesses an unused address, a reset is generated. 6.6 Sleep mode Entering the sleep mode will reset a part of the circuit. The reset is used to configure the circuit for correct wake-up after the sleep mode. If the SleepEn bit in the RegSysCtrl register has been set, the sleep mode can be entered by setting the bit Sleep in RegSysReset. During the sleep mode, the nresetsleep signal is active. For detailed information on the sleep mode, see the system documentation. 6.7 Control register description and operation Two registers are dedicated for reset status and control, RegSysReset and RegSysCtrl. The bits Sleep, SleepFlag and SleepEn are also located in those registers and are described in the chapter dedicated to the different operating modes of the circuit (system block). The RegSysReset register gives information on the source that generated the last reset. It can be read at the beginning of the application program to detect if the circuit is recovering from an error or exception condition, or if the circuit is starting up normally. • when ResetBusError is 1, a forbidden address access generated the reset. • when ResetWD is 1, the watchdog generated the reset. • when ResetfromPortA is 1, a PortA combination generated the reset. Note: If no bit is set to 1, the reset source was either the NRESET pin or the internal POR. Note: Several bits might be set or not, if the register was not cleared in between 2 reset occurrences. The two other bits concern the sleep mode control and information (see system documentation for the sleep mode description). • When SleepFlag is 1, the sleep mode was active before the reset occurred. This bit will always appear together with the ResetfromPortA bit since all other possibilities to leave the sleep mode (POR and NRESET pin) will clear the SleepFlag. • When Sleep is set to 1, and SleepEn is 1, the sleep mode is entered. The bit always reads back a 0. The RegSysCtrl register enables the different available reset sources and the sleep mode. • EnBusError enables the reset due to a bus error condition. • EnResetWD enables the reset due to the watchdog (can not be disabled once enabled). • EnResetPConf enables the reset of the port configurations when reset by Port A, a Bus Error or the watchdog. • SleepEn unlocks the Sleep bit. As long as SleepEn is 0, the Sleep bit has no effect. 6.8 Watchdog The watchdog is a timer, which has to be cleared at least every 2 seconds by the software to prevent a reset to be generated by the timeout condition. The watchdog can be enabled by software by setting the EnResetWD bit in the RegSysCtrl register to 1. It can then only be disabled by a power on reset or by setting the NRESET pin to a low state. The watchdog timer can be cleared by writing consecutively the values Hx0A and Hx03 to the RegSysWd register. The sequence must strictly be respected to clear the watchdog. © Semtech 2006 www.semtech.com 6-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver In assembler code, the sequence to clear the watchdog is: move AddrRegSysWd, #0x0A move AddrRegSysWd, #0x03 Only writing Hx0A followed by Hx03 resets the WD. If some other write instruction is done to the RegSysWd between the writing of the Hx0A and Hx03 values, the watchdog timer will not be cleared. It is possible to read the status of the watchdog in the RegSysWd register. The watchdog is a 4 bit counter with a count range between 0 and 7. The system reset is generated when the counter is reaching the value 8. 6.9 Start-up and watchdog specifications At start-up of the circuit, the POR block generates a reset signal during tPOR. The circuit starts software execution after this period (see system chapter). The POR is intended to force the circuit into a correct state at start-up. For precise monitoring of the supply voltage, the voltage level detector (VLD) has to be used. Symbol TPOR Parameter POR reset duration Vbat_sl Supply ramp up WDtime Watchdog timeout period Min Typ 5 Max Unit 20 ms Comments 0.5 V/ms 1 2 s 2 Table 6. Electrical and timing specifications Note: 1) The Vbat_sl defines the minimum slope required on VBAT. Correct start-up of the circuit is not guaranteed if this slope is too slow. In such a case, a delay has to be built using the NRESET pin. Note: 2) The minimal watchdog timeout period is guaranteed when the internal oscillators are used. In case an external clock source is used, the watchdog timeout period will be correct in so far the contents of the RegSysRCTrim1 and RegSysRCTrim2 registers are correct (see clock block documentation for more details). © Semtech 2006 www.semtech.com 6-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 7. Clock Generation 7.1 7.2 7.3 7.4 7.5 7.5.1 7.5.2 7.5.3 7.6 7.6.1 7.6.2 7.7 7.7.1 7.7.2 7.8 7.9 7.10 FEATURES ........................................................................................................................................... 7-2 OVERVIEW ........................................................................................................................................... 7-2 REGISTER MAP ..................................................................................................................................... 7-2 INTERRUPTS AND EVENTS MAP ............................................................................................................... 7-3 CLOCK SOURCES .................................................................................................................................. 7-5 RC oscillator Configuration................................................................................................................. 7-5 RC oscillator frequency tuning............................................................................................................ 7-5 RC oscillator specifications ................................................................................................................ 7-6 XTAL OSCILLATOR................................................................................................................................. 7-7 Xtal configuration ............................................................................................................................... 7-7 Xtal oscillator specifications................................................................................................................ 7-7 EXTERNAL CLOCK ................................................................................................................................. 7-8 External clock configuration ............................................................................................................... 7-8 External clock specification ................................................................................................................ 7-8 CLOCK SOURCE SELECTION ................................................................................................................... 7-8 PRESCALERS ....................................................................................................................................... 7-9 32 KHZ FREQUENCY SELECTOR ........................................................................................................... 7-10 © Semtech 2006 www.semtech.com 7-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 7.1 Features 3 available clock sources (RC oscillator, quartz oscillator and external clock). • 2 divider chains: high-prescaler (8 bits) and low-prescaler (15 bits). • CPU clock disabling in halt mode. 7.2 Overview The XE88LCxx chips can work on different clock sources (RC oscillator, quartz oscillator and external clock). The clock generator block is in charge of distributing the necessary clock frequencies to the circuit. Figure 7-1 represents the functionality of the clock block. The internal RC oscillator or an external clock source can be selected to drive the high prescaler. This prescaler generates frequency divisions down to 1/256 of its input frequency. A 32kHz clock is generated by enabling the quartz oscillator (if present in the product) or by selecting the appropriate tap on the high prescaler. The low prescaler generates clock signals from 32kHz down to 1Hz. The clock source for the CPU can be selected from the RC oscillator, the external clock or the 32kHz clock. 7.3 Register map pos. 7 6 5 4 3 2 1 0 RegSysClock CpuSel EnExtClock BiasRc ColdXtal EnableXtal EnableRc rw r/w r r/w r/w r r r/w r/w reset 0 nresetsleep 0 0 nresetcold 1 nresetcold 1 nresetsleep 0 0 nresetsleep 1 nresetsleep function Select speed for cpuck Unused Enable for external clock Enable Rcbias (reduces start-up time of RC). Xtal in start phase Unused Enable Xtal oscillator Enable RC oscillator Table 7-1: RegSysClock register pos. 7-2 1 0 RegSysMisc -Output16k OutputCpuCk rw r r/w r/w reset 000000 0 nresetsleep 0 nresetsleep function Unused Output 16 kHz signal on PB[3] Output CPU clock on PB[2] Table 7-2: RegSysMisc register pos. 7-1 0 RegSysPre0 -ClearLowPrescal rw r w1 r0 reset 0000000 0 function Unused Write 1 to reset low prescaler, but always reads 0 Table 7-3: RegSysPre0 register © Semtech 2006 www.semtech.com 7-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7-5 4 3 2 1 0 RegSysRcTrim1 -RcFreqRange RcFreqCoarse[3] RcFreqCoarse[2] RcFreqCoarse[1] RcFreqCoarse[0] rw r r/w r/w r/w r/w r/w reset 000 0 nresetcold 0 nresetcold 0 nresetcold 0 nresetcold 1 nresetcold function Unused Low/high freq. range (low=0) RC coarse trim bit 3 RC coarse trim bit 2 RC coarse trim bit 1 RC coarse trim bit 0 Table 7-4: RegSysRCTrim1 register pos. 7-6 5 4 3 2 1 0 RegSysRcTrim2 -RcFreqFine[5] RcFreqFine[4] RcFreqFine[3] RcFreqFine[2] RcFreqFine[1] RcFreqFine[0] rw r r/w r/w r/w r/w r/w r/w reset 00 0 nresetcold 0 nresetcold 0 nresetcold 0 nresetcold 0 nresetcold 0 nresetcold function Unused RC fine trim bit 5 RC fine trim bit 4 RC fine trim bit 3 RC fine trim bit 2 RC fine trim bit 1 RC fine trim bit 0 Table 7-5: RegSysRCTrim2 register pos. 7-1 0 RegSysPtckmode -Reserved rw r r/w reset 0000000 0 nresetglobal function Unused Reserved Table 7-6: RegSysPtckmode register 7.4 Interrupts and events map interrupt source ck128Hz ck1Hz Default mapping in the interrupt manager RegIrqHig(6) Default mapping in the event manager RegEvn(5) RegIrqMid(3) RegEvn(1) Table 7-7: Interrupts and events map © Semtech 2006 www.semtech.com 7-3 Xtal Ext RC by 2 div. 1 0 ckRCExt En Ext Clock high © Semtech 2006 7-4 0 1 1 0 system clock ck32kHz Cpu Sel 0 1 prescaler low EnableXtal and not(En Ext Clock) RegSysRcTrim1&2 EnableRc or En Ext Clock prescaler cpuck ck1Hz ck32kHz ... ckRCExt/256 ckRCExt ... XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Figure 7-1. Clock block structure www.semtech.com XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 7.5 7.5.1 Clock sources RC oscillator Configuration The RC oscillator is always turned on and selected for CPU and system operation at power-on reset, pad NRESET, and when exiting sleep mode. It can be turned off after the Xtal (quartz oscillator) has been started, after selection of the external clock or by entering sleep mode. The RC oscillator has two frequency ranges: sub-MHz (50 kHz to 0.5 MHz) and above-MHz (0.5 MHz to 5 MHz). Inside a range, the frequency can be tuned by software for coarse and fine adjustment. See registers RegSysRcTrim1 and RegSysRcTrim2. Bit EnableRc in register RegSysClock controls the propagation of the RC clock signal and the operation of the oscillator. The user can stop the RC oscillator by resetting the bit EnableRc. Entering the sleep mode disables the RC oscillator. Note: The RC oscillator bias can be maintained while the oscillator is disabled by setting the bit BiasRc in RegSysClock. This allows a faster restart of the RC oscillator at the cost of increased power consumption (see section 7.5.3). 7.5.2 RC oscillator frequency tuning The RC oscillator frequency can be set using the bits in the RegSysRcTrim1 and RegSysRcTrim2 registers. Figure 7-2 shows the nominal frequency of the RC oscillator as a function of these bits. The absolute value of the frequency for a given register content may change by ±35% from chip to chip due to the tolerances on the integrated capacitors and resistors. However, the modification of the frequency as a function of a modification of the register content is fairly precise. This means that the curves in Figure 7-2 can shift up and down but that the slope remains unchanged. The bit RcFreqRange modifies the oscillator frequency by a factor of 10. The upper curve in the figure corresponds to RcFreqRange=1. The RcFreqCoarse modifies the frequency of the oscillator by a factor (RcFreqCoarse+1). The figure represents the frequency for 5 different values of the bits RcFreqCoarse: for each value the frequency is multiplied by 2. Incrementing the RcFreqFine code, increases the frequency by about 1.4%. The frequency of the oscillator is therefor given by: fRC=fRcmin⋅(1+9⋅RcFreqRange)⋅(1+RcFreqCoarse)⋅(1.014)RcFreqFine with fRcmin the RC oscillator frequency if the registers are all 0. © Semtech 2006 www.semtech.com 7-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 1E+07 1 11 00 0 11 0 1 11 0 10 00 00 1 11 0 0 00 11 11 0 11 0 00 00 0 01 11 0 11 0 00 0 00 00 0 0 0 01 0 RcFreqFine(5:0) 0 0 0 10 0 00 11 11 0 00 11 11 0 00 0 10 00 0 00 0 00 Nominal RC oscillator frequency [Hz] 1 00 00 00 00 1 00 00 00 1 00 00 1 11 00 00 00 11 00 00 01 11 11 11 0 10 01 00 00 0 00 11 00 0 00 0 00 00 01 00 0 10 0 00 11 11 00 0 01 11 0 00 00 10 00 11 0 00 0 01 11 11 11 1E+05 11 00 11 10 11 11 00 10 01 11 0 11 10 11 0 11 0 10 RcFreqFine(5:0) 0 01 11 1E+06 00 01 1 11 RcFreqRange='1' RcFreqRange='0' 00 RcFreqCoarse(3:0) 1E+04 0000 0001 0011 0111 1111 Figure 7-2. RC oscillator nominal frequency tuning. 7.5.3 RC oscillator specifications sym fRCmin RcFreqFine RC_su description Lowest RC frequency fine tuning step startup time PSRR @ DC Supply voltage dependence Temperature dependence ∆f/∆T min 25 typ 40 1.4 30 3 TBD TBD 0.1 max 55 2.0 50 5 unit kHz % us us %/V %/V %/°C Comments Note 1 BiasRc=0 BiasRc=1 Note 2 Note 3 Table 7-8. RC oscillator specifications Note 1: this is the frequency tolerance when all trimming codes are 0. The frequency at start-up is about twice as high. Note 2: frequency shift as a function of VBAT with normal regulator function. Note 3: frequency shift as a function of VBAT while the regulator is short-circuited to VBAT. The tolerances on the minimal frequency and the drift with supply or temperature can be cancelled using the software or hardware DFLL (digital frequency locked loop) which uses the crystal oscillator as a reference frequency. © Semtech 2006 www.semtech.com 7-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 7.6 Xtal oscillator 7.6.1 Xtal configuration The Xtal operates with an external crystal of 32’768 Hz. During Xtal oscillator start-up, the first 32768 cycles are masked. The two bits EnableXtal and ColdXtal in register RegSysClock control the oscillator. At power-on reset, a pad NRESET pulse or during sleep mode, EnableXtal is reset and ColdXtal is set (Xtal oscillator is not selected at start-up). The user can start Xtal oscillator by setting EnableXtal. When the Xtal oscillator starts, bit ColdXtal is reset after 32768 cycles. Before ColdXtal is reset by the system, the Xtal frequency precision is not guaranteed. The Xtal oscillator can be stopped by the user by resetting bit EnableXtal. When the user enters into sleep mode, the Xtal is stopped. 7.6.2 Xtal oscillator specifications The crystal oscillator has been designed for a crystal with the specifications given in Table 7-9. The oscillator precision can only be guaranteed for this crystal. Symbol Fs CL Description Resonance frequency CL for nominal frequency Rm Cm C0 Rmp Motional resistance Motional capacitance Shunt capacitance Motional resistance of 6th overtone (parasitic) Quality factor Q Min Typ 32768 8.2 Max 1.8 0.7 4 40 2.5 1.1 8 100 3.2 2.0 kΩ fF pF kΩ 30k 50k 400k - 15 Unit Hz pF Comments Table 7-9. Crystal specifications. For safe operation, low power consumption and to meet the specified precision, careful board layout is required: Keep lines XIN and XOUT short and insert a VSS line in between them. Connect the crystal package to VSS. No noisy or digital lines near XIN or XOUT. Insert guards where needed. Respect the board specifications of Table 7-10. Symbol Rh_xin Rh_xout Rh_xin_xout Cp_xin Cp_xout Cp_xin_xout Description Resistance XIN-VSS Resistance XOUTVSS Resistance XINXOUT Capacitance XINVSS Capacitance XOUTVSS Capacitance XINXOUT Min 10 10 Typ Max 50 Unit MΩ MΩ Comments MΩ 0.5 3.0 pF 0.5 3.0 pF 0.2 1.0 pF Table 7-10. Board layout specifications. © Semtech 2006 www.semtech.com 7-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver The oscillator characteristics are given in Table 7-11. The characteristics are valid only if the crystal and board layout meet the specifications above. Symbol fXtal St_xtal Fstab Description Nominal frequency Start-up time Frequency deviation Min Typ 32768 1 -100 Max 2 300 Unit Hz s ppm Comments Note 1 Table 7-11. Crystal oscillator characteristics. Note 1. This gives the relative frequency deviation from nominal for a crystal with CL=8.2pF and within the temperature range -40°C to 85°C. The crystal tolerance, crystal aging and crystal temperature drift are not included in this figure. 7.7 External clock 7.7.1 External clock configuration The user can provide an external clock instead of the internal oscillators. The external provided frequency is internally divided by two. The external clock input pin is XIN. The system is configured for external clock by bit EnExtClock in register RegSysClock. Using the bits in the registers RegSysRcTrim1 and RegSysRcTrim2, the ck32kHz clock frequency can be controlled (see section 7.10). Note: when using the external clock, the Xtal is not available. 7.7.2 External clock specification The external clock has to satisfy the specifications in the table below. Correct behavior of the circuit can not be guaranteed if the external clock signal does not respect the specifications below. Symbol FEXT PW_1 PW_0 FEXT_LV PW_1_LV PW_0_LV Description External clock frequency Pulse 1 width Pulse 0 width External clock frequency Pulse 1 width Pulse 0 width Min Typ 0.06 0.03 TBD TBD Max 8 Unit MHz Comments Note 1 20 TBD µs µs kHz Note 1 Note 1 Note 2 µs µs Note 2 Note 2 20 Table 7-12. External clock specifications. Note 1. For VBAT≥2.4V Note 2. For VBAT=VREG=1.2V 7.8 Clock source selection There are three possible clock sources available for the CPU clock. The RC clock is always selected after powerup, a negative pulse on NRESET or after Sleep mode. The CPU clock selection is done with the bit CpuSel in RegSysClock (0= fastest clock, 1= 32 kHz from Xtal if EnableXtal =1 and EnExtClock = 0 else from high prescaler 32 kHz output). Switching from one clock source to another is glitch free. © Semtech 2006 www.semtech.com 7-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver The next table summarizes the different clock configurations of the circuit: EnableRc EnableXtal Sleep Xtal RC RC + Xtal External Clock targets EnExtClock Clock Sources Mode name 0 0 0 0 1 0 0 1 1 X 0 1 0 1 X Cpuck CpuSel=0 CpuSel=1 off Xtal RCNOTE 2 RCNOTE 1 and 2 External NOTE off Xtal High presc. Xtal High presc. 2 High Prescaler Clock input Low Prescaler Clock input off off RC RCNOTE 1 External off Xtal High presc. Xtal High presc. Table 7-13: Table of clocking modes. Note 1: The frequency of the RC must be higher than 100 kHz when Xtal is enabled in order to ensure a proper 32 kHz operation. Note 2: The clock RC can be divided by the value of freq instruction (see coolrisc instruction information) freq instruction nodiv div2 div4 div8 div16 cpuck RC or external RC/2 or external/2 RC/4 or external/4 RC/8 or external/8 RC/16 or external/16 Note 3: Switching from one clock source to another and stopping the unused clock source must be performed using 2 MOVE instructions to RegSysClock. First select the new clock source and then stop the unused one. 7.9 Prescalers The clock generator block embeds two divider chains: the high prescaler and the low one. The high prescaler is made of an 8 stage dividing chain and the low prescaler of a 15 stage dividing chain. Features: • High prescaler can only be driven with RC clock or external clock (bits EnableRc or EnExtClock have to be set, see Table 7-13). • Low prescaler can be driven from the high prescaler or directly with the Xtal clock when bit EnableXtal is set to 1 and bit EnExtClock is set to 0. • Bit ClearLowPrescal in the RegSysPre0 register allows to reset synchronously the low prescaler, the low prescaler is also automatically cleared when bit EnableXtal is set. Both dividing chains are reset asynchronously by the nresetglobal signal. • Bit ColdXtal=1 indicates the Xtal is in its start phase. It is active for 32768 Xtal cycles after setting EnableXtal. © Semtech 2006 www.semtech.com 7-9 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 7.10 32 kHz frequency selector A decoder is used to select from the high prescaler, the frequency tap that is the closest to 32 kHz to operate the low prescaler when the Xtal is not running. In this case, the RC oscillator frequency of ±35% will also be valid for the low prescaler frequency outputs. The next table shows how the RC trimming values in the RegSysRcTrim1 and RegSysRcTrim2 registers select the 32 kHz frequency. The least significant bits of the RcFreqFine word are not used. In order to ensure the correct frequency selection for the low prescaler when having an external clock, a proper value must be set in the RC trim registers. The code can be selected from the table below as a function of the frequency ratio between half the frequency of the external clock and 32kHz. If the frequency is not set correctly, all timings derived from the low prescaler will be shifted accordingly (e.g. watchdog frequencies) and some peripherals may no longer function correctly if the deviation from 32kHz is too large (e.g. the voltage level detector). © Semtech 2006 www.semtech.com 7-10 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver RcFreqRange&RcFreqCoarse(3:0)&RcFreqFine(5:3) Selected high prescaler tap Default case (0’0001’000) From 0’0000’000 to 0’0000’100 From 0’0000’101 to 0’0001’100 From 0’0001’101 to 0’0001’111 0’0010’000 From 0’0010’001 to 0’0010’110 0’0010’111 From 0’0011’000 to 0’0011’100 From 0’0011’101 to 0’0011’111 From 0’0100’000 to 0’1000’010 From 0’0100’011 to 0’0100’111 0’0101’000 From 0’0101’001 to 0’0101’110 0’0101’111 From 0’0110’000 to 0’0110’101 From 0’0110’110 to 0’0110’111 From 0’0111’000 to 0’0111’100 From 0’0111’101 to 0’0111’111 From 0’1000’000 to 0’1000’011 From 0’1000’100 to 0’1000’111 From 0’1001’000 to 0’1001’010 From 0’1001’011 to 0’1001’111 From 0’1010’000 to 0’1010’001 From 0’1010’010 to 0’1010’111 0’1011’000 From 0’1011’001 to 0’1011’110 0’1011’111 From 0’1100’000 to 0’1100’110 0’1100’111 From 0’1101’000 to 0’1101’101 From 0’1101’110 to 0’1101’111 From 0’1110’000 to 0’1110’100 From 0’1110’101 to 0’1110’111 From 0’1111’000 to 0’1111’100 From 0’1111’101 to 0’1111’111 From 1’0000’000 to 1’0000’010 From 1’0000’011 to 1’0001’010 From 1’0001’011 to 1’0010’100 From 1’0010’101 to 1’0010’111 From 1’0011’000 to 1’0011’010 From 1’0011’011 to 1’0011’111 1’0100’000 From 1’0100’001 to 1’0100’110 1’0100’111 From 1’0101’000 to 1’0101’100 From 1’0101’101 to 1’0101’111 From 10110’000 to 1’0110’011 From 1’0110’100 to 1’0110’111 From 1’0111’000 to 1’0111’010 From 1’0111’011 to 1’0111’111 From 1’1000’000 to 1’1000’001 From 1’1000’010 to 1’1000’111 1’1001’000 From 1’1001’001 to 1’1111’111 Ckrcext/2 Ckrcext Ckrcext/2 Ckrcext/4 Ckrcext/2 Ckrcext/4 Ckrcext/8 Ckrcext/4 Ckrcext/8 Ckrcext/4 Ckrcext/8 Ckrcext/4 Ckrcext/8 Ckrcext/16 Ckrcext/8 Ckrcext/16 Ckrcext/8 Ckrcext/16 Ckrcext/8 Ckrcext/16 Ckrcext/8 Ckrcext/16 Ckrcext/8 Ckrcext/16 Ckrcext/8 Ckrcext/16 Ckrcext/32 Ckrcext/16 Ckrcext/32 Ckrcext/16 Ckrcext/32 Ckrcext/16 Ckrcext/32 Ckrcext/16 Ckrcext/32 Ckrcext/8 Ckrcext/16 Ckrcext/32 Ckrcext/64 Ckrcext/32 Ckrcext/64 Ckrcext/32 Ckrcext/64 Ckrcext/128 Ckrcext/64 Ckrcext/128 Ckrcext/64 Ckrcext/128 Ckrcext/64 Ckrcext/128 Ckrcext/64 Ckrcext/128 Ckrcext/64 Ckrcext/128 Table 7-14: Table of 32kHz high prescaler tap decoder. © Semtech 2006 www.semtech.com 7-11 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 8. Interrupt Handler 8.1 8.2 8.3 8.4 8.5 FEATURES .............................................................................................................................8-2 OVERVIEW ............................................................................................................................8-2 REGISTER MAP ......................................................................................................................8-2 DETAILED DESCRIPTION ..........................................................................................................8-4 INTERRUPT HANDLING SOFTWARE............................................................................................8-5 © Semtech 2006 www.semtech.com 8-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 8.1 Features The XE8000 chips support 24 interrupt sources, divided into 3 levels of priority. 8.2 Overview The interrupt handler allows to manage 24 interrupt sources individually. The 24 interrupt sources are divided into 3 levels of priority: High (8 interrupt sources), Mid (8 interrupt sources), and Low (8 interrupt sources). Those 3 levels of priority are directly mapped to those supported by the CoolRisc (IN0, IN1 and IN2; see CoolRisc documentation for more information). Additional functions are given that allow fast detection of the highest priority interrupt that has been activated. 8.3 Register map Register name RegIrqHig RegIrqMid RegIrqLow RegIrqEnHig RegIrqEnMid RegIrqEnLow RegIrqPriority RegIrqIrq Table 8-1: IRQ handler registers pos. 7 RegIrqHig RegIrqHig[7] 6 RegIrqHig[6] 5 RegIrqHig[5] 4 RegIrqHig[4] 3 RegIrqHig[3] 2 RegIrqHig[2] 1 RegIrqHig[1] 0 RegIrqHig[0] rw r c1 r c1 r c1 r c1 r c1 r c1 r c1 r c1 reset 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function interrupt #23 (high priority) clear interrupt #23 when 1 is written interrupt #22 (high priority) clear interrupt #22 when 1 is written interrupt #21 (high priority) clear interrupt #21 when 1 is written interrupt #20 (high priority) clear interrupt #20 when 1 is written interrupt #19 (high priority) clear interrupt #19 when 1 is written interrupt #18 (high priority) clear interrupt #18 when 1 is written interrupt #17 (high priority) clear interrupt #17 when 1 is written interrupt #16 (high priority) clear interrupt #16 when 1 is written Table 8-2: RegIrqHig © Semtech 2006 www.semtech.com 8-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7 RegIrqMid RegIrqMid[7] 6 RegIrqMid[6] 5 RegIrqMid[5] 4 RegIrqMid[4] 3 RegIrqMid[3] 2 RegIrqMid[2] 1 RegIrqMid[1] 0 RegIrqMid[0] rw r c1 r c1 r c1 r c1 r c1 r c1 r c1 r c1 reset 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function interrupt #15 (mid priority) clear interrupt #15 when 1 is written interrupt #14 (mid priority) clear interrupt #14 when 1 is written interrupt #13 (mid priority) clear interrupt #13 when 1 is written interrupt #12 (mid priority) clear interrupt #12 when 1 is written interrupt #11 (mid priority) clear interrupt #11 when 1 is written interrupt #10 (mid priority) clear interrupt #10 when 1 is written interrupt #9 (mid priority) clear interrupt #9 when 1 is written interrupt #8 (mid priority) clear interrupt #8 when 1 is written Table 8-3: RegIrqMid pos. 7 RegIrqLow RegIrqLow[7] 6 RegIrqLow[6] 5 RegIrqLow[5] 4 RegIrqLow[4] 3 RegIrqLow[3] 2 RegIrqLow[2] 1 RegIrqLow[1] 0 RegIrqLow[0] rw r c1 r c1 r c1 r c1 r c1 r c1 r c1 r c1 reset 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function interrupt #7 (low priority) clear interrupt #7 when 1 is written interrupt #6 (low priority) clear interrupt #6 when 1 is written interrupt #5 (low priority) clear interrupt #5 when 1 is written interrupt #4 (low priority) clear interrupt #4 when 1 is written interrupt #3 (low priority) clear interrupt #3 when 1 is written interrupt #2 (low priority) clear interrupt #2 when 1 is written interrupt #1 (low priority) clear interrupt #1 when 1 is written interrupt #0 (low priority) clear interrupt #0 when 1 is written Table 8-4: RegIrqLow pos. 7 6 5 4 3 2 1 0 RegIrqEnHig RegIrqEnHig[7] RegIrqEnHig[6] RegIrqEnHig[5] RegIrqEnHig[4] RegIrqEnHig[3] RegIrqEnHig[2] RegIrqEnHig[1] RegIrqEnHig[0] rw rw rw rw rw rw rw rw rw reset 0 0 0 0 0 0 0 0 function 1= enable interrupt #23 1= enable interrupt #22 1= enable interrupt #21 1= enable interrupt #20 1= enable interrupt #19 1= enable interrupt #18 1= enable interrupt #17 1= enable interrupt #16 Table 8-5: RegIrqEnHig © Semtech 2006 www.semtech.com 8-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7 6 5 4 3 2 1 0 RegIrqEnMid RegIrqEnMid[7] RegIrqEnMid[6] RegIrqEnMid[5] RegIrqEnMid[4] RegIrqEnMid[3] RegIrqEnMid[2] RegIrqEnMid[1] RegIrqEnMid[0] rw rw rw rw rw rw rw rw rw reset 0 0 0 0 0 0 0 0 function 1= enable interrupt #15 1= enable interrupt #14 1= enable interrupt #13 1= enable interrupt #12 1= enable interrupt #11 1= enable interrupt #10 1= enable interrupt #9 1= enable interrupt #8 Table 8-6: RegIrqEnMid pos. 7 6 5 4 3 2 1 0 RegIrqEnLow RegIrqEnLow[7] RegIrqEnLow[6] RegIrqEnLow[5] RegIrqEnLow[4] RegIrqEnLow[3] RegIrqEnLow[2] RegIrqEnLow[1] RegIrqEnLow[0] rw rw rw rw rw rw rw rw rw reset 0 0 0 0 0 0 0 0 function 1= enable interrupt #7 1= enable interrupt #6 1= enable interrupt #5 1= enable interrupt #4 1= enable interrupt #3 1= enable interrupt #2 1= enable interrupt #1 1= enable interrupt #0 Table 8-7: RegIrqEnLow pos. 7-0 RegIrqPriority RegIrqPriority rw r reset 11111111 function code of highest priority set Table 8-8: RegIrqPriority pos. 7-3 2 1 0 RegIrqIrq IrqHig IrqMid IrqLow rw r r r r reset 00000 0 0 0 function unused one or more high priority interrupts is set one or more mid priority interrupts is set one or more low priority interrupts is set Table 8-9: RegIrqIrq 8.4 Detailed description The CoolRISC core has 3 different interrupt levels IN0, IN1 and IN2 (Figure 8-1). When these interrupts are triggered, the program counter (PC) is loaded with a fixed address. In case more than one interrupt occurs simultaneously, the execution order is IN0, IN1, IN2. The masking, setting and clearing of these interrupts can be done in the stat register (see chapter describing the CPU). The interrupt handler bundles a certain number of interrupt sources and routes them to one of these three interrupts and provides the possibility to enable/disable each of them individually. The definition of the interrupt sources is given in the memory mapping chapter. RegIrqHig, RegIrqMid, and RegIrqLow are 8-bit registers containing flags for the interrupt sources. Those flags are set when the interrupt is enabled (i.e. if the corresponding bit in the registers RegIrqEnHig, RegIrqEnMid or RegIrqEnLow is set) and a rising edge is detected on the corresponding interrupt source. © Semtech 2006 www.semtech.com 8-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Once memorized, an interrupt flag can be cleared by writing a ‘1’ in the corresponding bit of RegIrqHig, RegIrqMid or RegIrqLow. Writing a ‘0’ does not modify the flag. To definitively clear the interrupt, one has to clear the CoolRISC interrupt in the CoolRISC stat register. All interrupts are automatically cleared after a reset. Two registers are provided to facilitate the writing of interrupt service software. RegIrqPriority contains the number of the highest priority set (its value is 0xFF when no interrupt is memorized). RegIrqIrq indicates the priority level of the currently activated interrupts. stat IE2 IE1 GIE IN2 … high priority PC=h0003 medium priority PC=h0001 low priority PC=h0002 All interrupt sources are sampled by the highest frequency in the system. A CPU interruption is generated and memorized when an interrupt becomes high. Between the rising edge of the interrupt on the peripheral and the rising edge on the CoolRISC core, there is a latency of one clock cycle. IN1 IN0 EV1 EV0 … RegIrqLow RegIrqMid 7 6 5 4 3 2 1 0 RegIrqEnHig 7 6 5 4 3 2 1 0 interrupt sources RegIrqHig Figure 8-1. Principle of the interrupt handler. 8.5 Interrupt handling software This chapter describes an example of the software used for the interrupt handler. This software is present by default in the software development environments. It represents only one of several possible ways of handling the interrupts. © Semtech 2006 www.semtech.com 8-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver First of all, the jump addresses are defined at the beginning of the crt0.s file. In our case, all three interrupt levels jump to the same place (defined by the _interrupt label), but this can be changed if required. ######################################################################## ## Reset & interrupt vectors ######################################################################## _start: jump main_init ; reset jump _interrupt ; IN1 jump _interrupt ; IN2 jump _interrupt ; IN0 The first thing to do when an interrupt is activated is to save the context. You have to start with saving the contents of the accumulator, then the flags and finally the internal CPU registers. You will find this part of the code in the IRQComon_xx.s file. _interrupt: ######################################################################## ## Save all registers and flags ######################################################################## move -(i3), a move a, r0 sflag move -(i3), a move -(i3), ipl move -(i3), iph move -(i3), i0l move -(i3), i0h move -(i3), i1l move -(i3), i1h move -(i3), i2l move -(i3), i2h move -(i3), r0 move -(i3), r1 move -(i3), r2 move -(i3), r3 Next step is to determine which interrupt is activated. In this case, we use the value in the RegIrqPriority register to determine the highest priority interrupt that was activated. Other ways can be used, especially when the priority order fixed in the hardware needs to be changed. You will find this part of the code in the IRQComon_xx.s file. In this example, the labels are used as defined for the XE8802. ######################################################################## ## The following lines enables the adress calculation of the interrupt ## table. Where RegIrqPriority is the addres offset for the table. ## The RegIrqPriority valid values are between 0x00 until 0x017. The ## 0xFF value should never exist. ######################################################################## move r0,RegIrqPriority calls _interrupttab ; save pc+1 in ip _interrupttab: add ipl,#0x05 ; add the offset, nb instr. before table addc iph,#0x00 ; propagate carry add ipl,r0 ; add the offset of the regirqpriority addc iph,#0x00 ; propagate carry rets ; put ip in pc ; interrupt table jump ret_int jump ret_int jump Irq_Pa2 jump Irq_Pa3 jump Irq_CntD jump Irq_CntB jump Irq_Pa6 jump Irq_Pa7 jump Irq_Pa0 ; ; ; ; ; ; ; ; ; RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority © Semtech 2006 = = = = = = = = = 0x00 0x01 0x02 0x03 0x04 0x05 0x06 0x07 0x08 www.semtech.com 8-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver jump jump jump jump jump jump jump jump jump jump jump jump jump jump jump Irq_Pa1 Irq_Vld Irq_1Hz Irq_Pa4 Irq_Pa5 Irq_UsrtCond1 Irq_UsrtCond2 Irq_UartRx Irq_UartTx Irq_Cmpd Irq_CntC Irq_CntA Irq_Spi Irq_128Hz Irq_AC ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority RegIrqPriority = = = = = = = = = = = = = = = 0x09 0x0A 0x0B 0x0C 0x0D 0x0E 0x0F 0x10 0x11 0x12 0x13 0x14 0x15 0x16 0x17 The next steps are to clear the interrupt flag in the interrupt handler, to call the specific function for the identified interrupt source and to clear the interrupt in the stat register. This code can be found in the file IRQSave0_xx.s. ######################################################################## Irq_AC: move RegIrqHig, #0x80 calls Handle_Irq_AC jump ret_int0 ######################################################################## Irq_128Hz: move RegIrqHig, #0x40 calls Handle_Irq_128Hz jump ret_int0 ######################################################################## Irq_Spi: move RegIrqHig, #0x20 calls Handle_Irq_Spi jump ret_int0 … ret_int0: clrb stat, #2 jump ret_int Finally, the context and the PC have to be restored. This code can be found in the IRQComon_xx.s file. ret_int: ######################################################################## ## Restore all registers and flags ######################################################################## move r3, (i3)+ move r2, (i3)+ move r1, (i3)+ move r0, (i3)+ move i2h, (i3)+ move i2l, (i3)+ move i1h, (i3)+ move i1l, (i3)+ move i0h, (i3)+ move i0l, (i3)+ move iph, (i3)+ move ipl, (i3)+ rflag (i3)+ move a, (i3)+ reti ######################################################################## ## End of interrupt handlers ######################################################################## © Semtech 2006 www.semtech.com 8-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 9. Event Handler 9.1 9.2 9.3 9.4 FEATURES .............................................................................................................................9-2 OVERVIEW ............................................................................................................................9-2 REGISTER MAP ......................................................................................................................9-2 DETAILED DESCRIPTION ..........................................................................................................9-3 © Semtech 2006 www.semtech.com 9-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 9.1 Features The XE8000 chips support 8 event sources, divided into 2 levels of priority. 9.2 Overview An event is different from an interrupt in that it does not modify the program counter (PC). Events are used by two microcontroller instructions. First of all, events are useful to wake-up the microcontroller when it is in HALT. The software execution then simply resumes at the instruction next to the HALT instruction. The second instruction is the conditional jump on event (JEV). The jump is executed if one of the event flags in the stat register is set. In all other cases, the occurrence of an event has no effect. The event handler allows to manage 8 event sources individually. The 8 event sources are divided into 2 levels of priority: High (4 event sources) and Low (4 event sources). Those 2 levels of priority are directly mapped to those supported by the CoolRisc™ (EV0and IN1; see CoolRisc™ documentation for more information). Additional functions are given that allow fast detection of the highest priority event that has been activated. 9.3 Register map The addresses given in Table 9-1 are the default values and may be different in some products. Register name RegEvn RegEvnEn RegEvnPriority RegEvnEvn Table 9-1: EVN handler registers. pos. 7 RegEvn RegEvn[7] 6 RegEvn[6] 5 RegEvn[5] 4 RegEvn[4] 3 RegEvn[3] 2 RegEvn[2] 1 RegEvn[1] 0 RegEvn[0] rw r c1 r c1 r c1 r c1 r c1 r c1 r c1 r c1 reset 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function event #7 (high priority) clear event #7 when written 1 event #6 (high priority) clear event #6 when written 1 event #5 (high priority) clear event #5 when written 1 event #4 (high priority) clear event #4 when written 1 event #3 (low priority) clear event #3 when written 1 event #2 (low priority) clear event #2 when written 1 event #1 (low priority) clear event #1 when written 1 event #0 (low priority) clear event #0 when written 1 Table 9-2: RegEvn 9-2 D0405-134 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7 6 5 4 3 2 1 0 RegEvnEn RegEvnEn[7] RegEvnEn[6] RegEvnEn[5] RegEvnEn[4] RegEvnEn[3] RegEvnEn[2] RegEvnEn[1] RegEvnEn[0] rw rw rw rw rw rw rw rw rw reset 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function 1= enable event #7 1= enable event #6 1= enable event #5 1= enable event #4 1= enable event #3 1= enable event #2 1= enable event #1 1= enable event #0 Table 9-3: RegEvnEn pos. 7-0 RegEvnPriority RegEvnPriority rw r reset 11111111 nresetglobal function code of highest event set FF if no event present. Table 9-4: RegEvnPriority pos. 7-2 1 RegEvnEvn EvnHig rw r r reset 00000 0 nresetglobal 0 EvnLow r 0 nresetglobal function unused one or more high priority event is set one or more low priority event is set Table 9-5: RegEvnEvn 9.4 Detailed description The CoolRISC core has 2 different event levels EV0 and EV1 (Figure 9-1). The setting and clearing of these events can be done in the stat register (see chapter describing the CPU). The event handler bundles a certain number of event sources and routes them to one of these two events and provides the possibility to enable/disable each of them individually. The definition of the event sources is given in the memory mapping chapter. RegEvn is an 8-bit register containing flags for the event sources. Those flags are set when the event is enabled (i.e. if the corresponding bit in the registers RegEvnEn is set) and a rising edge is detected on the corresponding event source. Once memorized, writing a ‘1’ in the corresponding bit of RegEvn clears an event flag. Writing a ‘0’ does not modify the flag. All interrupts are automatically cleared after a reset. Two registers are provided to facilitate the writing of event service software. RegEvnPriority contains the number of the highest event set (its value is 0xFF when no event is memorized). RegEvnEvn indicates the priority level of the current events. All event sources are sampled by the highest frequency in the system. A CPU event is generated and memorized when an event becomes high. The 8 event sources are divided into 2 levels of priority: High (4 event sources) and Low (4 event sources). Those 2 levels of priority are directly mapped to those supported by the CoolRisc (EV0 and EV1; see CoolRisc documentation for more information). 9-3 D0405-134 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver IE2 IE1 GIE IN2 IN1 IN0 EV1 EV0 RegEvn 7 6 5 4 3 2 1 0 RegEvnEn 7 6 5 4 3 2 1 0 event sources stat Figure 9-1. Event handler principle. 9-4 D0405-134 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 10. 10.1 10.1.1 10.2 Low Power RAM FEATURES .......................................................................................................................................... 10-2 Overview.......................................................................................................................................... 10-2 REGISTER MAP ................................................................................................................................... 10-2 © Semtech 2006 www.semtech.com 10-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 10.1 Features 10.1.1 Overview In order to save power consumption, 8 8-bit registers are provided in page 0. These memory locations should be reserved for often-updated variables. As they are real registers and not RAM, power consumption is greatly reduced. 10.2 Register map pos. 7-0 7-0 7-0 7-0 7-0 7-0 7-0 7-0 Reg00 Reg00 Reg01 Reg02 Reg03 Reg04 Reg05 Reg06 Reg07 rw rw rw rw rw rw rw rw rw reset 0 0 0 0 0 0 0 0 function low-power data memory low-power data memory low-power data memory low-power data memory low-power data memory low-power data memory low-power data memory low-power data memory Table 10-1: Low Power RAM © Semtech 2006 www.semtech.com 10-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 11. 11.1 11.2 11.3 11.4 11.5 11.6 Port A FEATURES .................................................................................................................................... 11-2 OVERVIEW .................................................................................................................................... 11-2 REGISTER MAP .............................................................................................................................. 11-3 INTERRUPTS AND EVENTS MAP........................................................................................................ 11-4 PORT A (PA) OPERATION .............................................................................................................. 11-4 PORT A ELECTRICAL SPECIFICATION ............................................................................................... 11-6 © Semtech 2006 www.semtech.com 11-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 11.1 Features • • • • • • • • input port, 8 bits wide each bit can be set individually for debounced or direct input each bit can be set individually for pullup or not snap-to-rail option for each input each bit is an interrupt request source on the rising or falling edge a system reset can be generated on an input pattern PA[0] and PA[1] can generate two events for the CPU, individually maskable PA[0] to PA[3] can be used as clock inputs for the counters/timers/PWM (product dependent) 11.2 Overview PortA is a general purpose 8 bit wide digital input port, with interrupt capability. Figure 11-1 shows its structure. Port A VBat 8 logic RegPASnapToRail 8 RegPAPullup 8 RegPADebounce 8x debounce 8 0 RegPACtrl 8 8 1 RegPAIn RegPAEdge 1 0 DebFast (RegPACtrl(0)) 1 8 events 8x cntclocks 1kHz 32kHz 8 8 Vss interrupts 0 RegPARes1 RegPARes0 1 11 10 01 0 resetfromporta 00 8x Figure 11-1: structure of Port A © Semtech 2006 www.semtech.com 11-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 11.3 Register map There are eight registers in Port A (PA), namely RegPAIn, RegPADebounce, RegPAPullup, RegPAEdge, RegPARes0, RegPARes1, RegPACtrl and RegPASnaptorail. Table 11-2 to Table 11-9 show the mapping of control bits and functionality of these registers register name RegPAIn RegPADebounce RegPAEdge RegPAPullup RegPARes0 RegPARes1 RegPACtrl RegPASnaptorail Table 11-1: PA registers pos. 7:0 RegPAIn PAIn[7:0] rw r reset description pad PA[7] to PA[0] input value x Table 11-2: RegPAIn pos. 7:0 RegPADebounce PADebounce[7:0] rw reset description rw 0 nresetpconf PA[7] to PA[0] 1: debounce enabled 0: debounce disabled Table 11-3: RegPADebounce pos. 7:0 RegPAEdge PAEdge[7:0] rw reset rw 0 nresetglobal description PA[7] to PA[0] edge configuration 0: positive edge 1: negative edge Table 11-4: RegPAEdge pos. 7:0 RegPAPullup PAPullup[7:0] rw reset rw 1 nresetpconf description PA[7] to PA[0] pullup enable 0: pullup disabled 1: pullup enabled Table 11-5: RegPAPullup pos. 7:0 RegPARes0 PARes0[7:0] rw rw reset 0 nresetglobal description PA[7] to PA[0] reset configuration Table 11-6: RegPARes0 pos. 7:0 RegPARes1 PARes1[7:0] rw rw reset 0 nresetglobal description PA[7] to PA[0] reset configuration Table 11-7: RegPARes © Semtech 2006 www.semtech.com 11-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7:1 0 RegPACtrl [7:1] DebFast rw r rw reset 0000000 0 nresetpconf description Unused 0 = slow debounce, 1 = fastdebounce Table 11-8: RegPACtrl pos. 7:0 RegPASnaptorail PASnaptorail[7:0] rw rw reset 0 nresetpconf description set snap-to-rail input on Table 11-9: RegPASnaptorail Note: Depending on the status of the EnResetPConf bit in RegSysCtrl, RegPAEdge, RegPADebounce and RegPACtrl can be reset by any of the possible system resets or only with power-on reset and NRESET pad. 11.4 Interrupts and events map Interrupt source pa_irqbus[5] pa_irqbus[4] pa_irqbus[1] pa_irqbus[0] pa_irqbus[7] pa_irqbus[6] pa_irqbus[3] pa_irqbus[2] Default mapping in the interrupt manager RegIrqMid[5] RegIrqMid[4] RegIrqMid[1] RegIrqMid[0] RegIrqLow[7] RegIrqLow[6] RegIrqLow[3] RegIrqLow[2] Default mapping in the event manager RegEvn[4] RegEvn[0] 11.5 Port A (PA) Operation The Port A input status (debounced or not) can be read from RegPAin. Debounce mode: Each bit in Port A can be individually debounced by setting the corresponding bit in RegPADebounce. After reset, the debounce function is disabled. After enabling the debouncer, the change of the input value is accepted only if height consecutive samples are identical. Selection of the clock is done by bit DebFast in Register RegPACtrl. DebFast Clock filter 0 1kHz 1 32kHz Table 10: debounce frequency selection Note: The tolerance on the debounce frequency depends on the selected clock source. When the external clock is used, the pulse width will be correct if the input of the low prescaler is set to a frequency close to 32kHz (see clock block documentation). Pullups/Snap-to-rail: Different functions are possible depending on the value of the registers RegPAPullup and RegPASnaptorail. When the corresponding bit in RegPAPullup is set to 0, the inputs are floating (pullup and pulldown resistors are disconnected). When the corresponding bit in RegPAPullup is 1 and in RegPASnaptorail is 0, a pullup resistor is connected to the input pin. Finally, when the corresponding bit in RegPAPullup is 1 and in RegPASnaptorail is 1, the snap-to-rail function is active. © Semtech 2006 www.semtech.com 11-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver The snap-to-rail function connects a pullup or pulldown resistor to the input pin depending on the value forced on the input pin. This function can be used for instance when the input port is connected to a tristate bus. When the bus is floating, the pullup or pulldown maintains the bus in the last low impedance state before it became floating until another low impedance output drives the bus. It also reduces the power consumption with respect to a classic pullup since it selects the pullup or pulldown resistor so that it confirms the detected input state. The state of input pin is summarized in the table below. PAPullup[x] PASnaptorail[x] 0 1 1 1 x 0 1 1 (last) externally forced PA[x] value x x 0 1 PA[x] pull floating pullup pulldown pullup Table 11: Snap-to-rail Port A starts up with the pullup resistor connected and the snap-to-rail function disabled. Port A as an interrupt source: Each Port A input is an interrupt request source and can be set on rising or falling edge with the corresponding bit in RegPAEdge. After reset, the rising edge is selected for interrupt generation by default. The interrupt source can be debounced by setting register RegPADebounce. The interrupt signals are sampled on the fastest clock in the circuit. In order to guarantee that the circuit detects the interrupt, the minimal pulse length should be 1 cycle of this clock. Note: care must be taken when modifying RegPAEdge because this register performs an edge selection. The change of this register may result in a transition, which may be interpreted as a valid interruption. Port A as an event source: The interrupt signals of the pins PA[0] and PA[1] are also available as events on the event controller. Port A as a clock source (product dependent): Images of the PA[0] to PA[3] input ports (debounced or not) are available as clock sources for the counter/timer/PWM peripheral. Port A as a reset source: Port A can be used to generate a system reset by placing a predetermined word on Port A externally. The reset is built using a logical and of the 8 PARes[x] signals: resetfromportA = PAReset[7] AND PAReset[6] AND PAReset[5] AND ... AND PAReset[0] PAReset[x] is itself a logical function of the corresponding pin PA[x]. One of four logical functions can be selected for each pin by writing into two registers RegPARes0 and RegPARes1 as shown in Table 11-12. PARes1[x] 0 0 1 1 PARes0[x] 0 1 0 1 PAReset[x] 0 PA[x] not(PA[x]) 1 Table 11-12: Selection bits for reset signal A reset from Port A can be inhibited by placing a 0 on both PARes1[x] and PARes0[x] for at least 1 pin. Setting both PARes1[x] and PARes0[x] to 1, makes the reset independent of the value on the corresponding pin. Setting both registers to hFF, will reset the circuit independent from the Port A input value. This makes it possible to do a reset by software. © Semtech 2006 www.semtech.com 11-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Note: depending of the value of PA[0] to PA[7], changes to RegPARes0 and RegPARes1 can cause a reset. Therefore it is safe to have always one (RegPARes0[x], RegPARes1[x]) equal to ‘00’ during the setting operations. 11.6 Port A electrical specification sym VINH VINL RPU Cin description Input high voltage Input low voltage Pull-up resistance Input capacitance min 0.7*VBAT VSS 20 typ 50 2.5 max VBAT 0.2*VBAT 80 unit Comments V VBAT≥2.4V V VBAT≥2.4V kΩ pF Note 1 Note 1: this value is indicative only since it depends on the package. Table 11-13. Electrical specification © Semtech 2006 www.semtech.com 11-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 12. Port B 12.1 12.2 12.3 12.4 12.5 12.5.1 12.5.2 12.6 12.7 12.7.1 12.7.2 12.8 FEATURES ......................................................................................................................................... 12-2 OVERVIEW ......................................................................................................................................... 12-2 REGISTER MAP ................................................................................................................................... 12-2 PORT B CAPABILITIES ......................................................................................................................... 12-3 PORT B ANALOG CAPABILITY................................................................................................................ 12-3 Port B analog configuration .............................................................................................................. 12-3 Port B analog function specification ................................................................................................. 12-5 PORT B FUNCTION CAPABILITY ............................................................................................................. 12-5 PORT B DIGITAL CAPABILITIES .............................................................................................................. 12-6 Port B digital configuration ............................................................................................................... 12-6 Port B digital function specification................................................................................................... 12-7 LOW POWER COMPARATORS ............................................................................................................... 12-7 © Semtech 2006 www.semtech.com 2-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 12.1 Features • • • • • • • Input / output / analog port, 8 bits wide Each bit can be set individually for input or output Each bit can be set individually for open-drain or push-pull Each bit can be set individually for pull-up or not (for input or open-drain mode) In open-drain mode, pull-up is not active when corresponding pad is set to zero The 8 pads can be connected individually to four internal analog lines (4 line analog bus) Two internal freq. (16 kHz and cpuck) can be output on PB[2] and PB[3] Product dependant: • Two PWM signal can be output on pads PB[0] and PB[1] • The synchronous serial interface (USRT) uses pads PB[5], PB[4] • The UART interface uses pads PB[6] and PB[7] for Tx and Rx 12.2 Overview Port B is a multi-purpose 8 bit Input/output port. In addition to digital behavior, all pins can be used for analog signals. Each port terminal can be individually selected as digital input or output or as analog for sharing one of four possible analog lines. 12.3 Register map Table 12-1 shows the Port B registers. register name RegPBOut RegPBIn RegPBDir RegPBOpen RegPBPullup RegPBAna Table 12-1: Port B registers pos. 7–0 RegPBOut PBOut[7-0] rw rw reset 0 nresetpconf description in digital mode Pad PB[7-0] output value description in analog mode Analog bus selection for pad PB[7-0] Table 12-2: RegPBOut RegPBIn pos. 7–0 PBIn[7-0] rw rw reset X description in digital mode Pad PB[7-0] input status description in analog mode Unused Table 12-3: RegPBIn RegPBDir pos. 7–0 PBDir [7-0] rw rw reset 0 nresetpconf description in digital mode Pad PB[7-0] direction (0=input) description in analog mode Analog bus selection for pad PB[7-0] Table 12-4: RegPBDir © Semtech 2006 www.semtech.com 12-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7–0 RegPBOpen PBOpen[7-0] rw rw reset description in digital mode 0 nresetpconf Pad PB[7-0] open drain (1 = open drain) description in analog mode Unused Table 12-5: RegPBOpen pos. 7 –0 RegPBPullup PBPullup[7] rw rw reset description in digital mode 1 nresetpconf Pull-up for pad PB[7-0] (1=active) description in analog mode Connect pad PB[7-0] on selected ana bus Table 12-6: RegPBPullup pos. 7–0 RegPBAna PBAna [7-0] rw rw reset description in digital mode 0 nresetpconf Set PB[7-0] in analog mode description in analog mode Set PB[7-0] in analog mode Table 12-7: RegPBAna Note: Depending on the status of the EnResPConf bit in RegSysCtrl, the reset conditions of the registers are different. See the reset block documentation for more details on the nresetpconf signal. 12.4 Port B capabilities Port B name PB[7] PB[6] PB[5] PB[4] PB[3] PB[2] PB[1] PB[0] high (analog) analog analog analog analog analog analog analog analog utilization (priority) medium (functions) uart Rx uart Tx usrt S1 usrt S0 16 kHz clock CPU PWM1 Counter C (C+D) PWM0 Counter A (A+B) low (digital) (default) I/O (with pull-up) I/O (with pull-up) I/O (with pull-up) I/O (with pull-up) I/O (with pull-up) I/O (with pull-up) I/O (with pull-up) I/O (with pull-up) Table 12-8: Different Port B functions Table 12-8 shows the different usages that can be made of the port B with the order of priority. If a pin is selected to be analog, it overwrites the function and digital set-up. If the pin is not selected as analog, but a function is enabled, it overwrites the digital set-up. If neither the analog nor function is selected for a pin, it is used as an ordinary digital I/O. This is the default configuration at start-up. Note: the presence of the functions is product dependent. 12.5 Port B analog capability 12.5.1 Port B analog configuration Port B terminals can be attached to a 4 line analog bus by setting the PBAna[x] bits to 1 in the RegPBAna register. The other registers then define the connection of these 4 analog lines to the different pads of Port B. These can be used to implement a simple LCD driver or A/D converter. Analog switching is available only when the circuit is © Semtech 2006 www.semtech.com 12-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver powered with sufficient voltage (see specification below). Below the specified supply voltage, only voltages that are close to VSS or VBAT can be switched. When PBAna[x] is set to 1, one pad of the Port B terminals is changed from digital I/O mode to analog. The usage of the registers RegPBPullup, RegPBOut and RegPBDir define the analog configuration (see Table 12-9). When PBAna[x] = 1, then PBPullup[x] connects the pin to the analog bus. PBDir[x] and PBPOut[x] select which of the 4 analog lines is used. analog bus selection PBDir[x] 0 0 1 1 X PBPullup[x] PB[x] selection on 1 1 1 1 0 analog line 0 analog line 1 analog line 2 analog line 3 High impedance PBout[x] 0 1 0 1 X Table 12-9: Selection of the analog lines with RegPBDir, RegPBout and RegPBPullup when PBAna[x] = 1 Example: Set the pads PB[2] and PB[5] on the analog line 3. (the values X depend on the configuration of others pads) • apply high impedance in the analog mode (move RegPBPullup,#0bXX0XX0XX) • go to analog mode (move RegPBAna,#0bXX1XX1XX) • select the analog line3 (move RegPBDir,#0bXX1XX1XX and move RegPBOut,#0bXX1XX1XX) • apply the analog line to the output (move RegPBPullup,#0bXX1XX1XX) © Semtech 2006 www.semtech.com 12-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 12.5.2 Port B analog function specification The table below defines the on-resistance of the switches between the pin and the analog bus for different conditions. The series resistance between 2 pins of Port B connected to the same analog line is twice the resistance given in the table. sym Ron Ron Cin Cin description switch resistance switch resistance input capacitance (off) input capacitance (on) min typ 3.5 4.5 max 11 15 unit kΩ kΩ pF pF Comments Note 1 Note 2 Note 3 Note 4 Table 12-10. Analog input specifications. Note 1: This is the series resistance between the pad and the analog line in 2 cases 1. VBAT ≥ 2.4V and the VMULT peripheral is present on the circuit and enabled. 2. VBAT ≥ 3.0V and the VMULT peripheral is not present on the circuit. Note 2: This is the series resistance in case VBAT ≥ 2.8V and the peripheral VMULT is not present on the circuit. Note 3: This is the input capacitance seen on the pin when the pin is not connected to an analog line. This value is indicative only since it is product and package dependent. Note 4: This is the input capacitance seen on the pin when the pin is connected to an analog line and no other pin is connected to the same analog line. This value is indicative only since it is product and package dependent. 12.6 Port B function capability The Port B can be used for different functions implemented by other peripherals. The description below is applicable only in so far the circuit contains these peripherals. When the counters are used to implement a PWM function (see the documentation of the counters), the PB[0] and PB[1] terminals are used as outputs (PB[0] is used if CntPWM0 in RegCntConfig1 is set to 1, PB[1] is used if CntPWM1 in RegCntConfig1 is set to 1) and the PWM generated values override the values written in RegPBout. However, PBDir(0) and PBDir(1) are not automatically overwritten and have to be set to 1. If Output16k is set in RegSysMisc, the frequency is output on PB[3]. This overrides the value contained in PBOut(3). However, PBDir(3) must be set to 1. The frequency and duty cycle of the clock signal are given in Figure 12-1. fmax is the frequency of fastest clock present in the circuit. 1/fmax 1/16k Figure 12-1. 16 kHz output clock timing Similarly, if OutputCkCpu is set in RegSysMisc, the CPU frequency is output on PB[2]. This overrides the value contained in PBOut(2). However, PBDir(2) must be set to 1. © Semtech 2006 www.semtech.com 12-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 1/f1 1/f2 Figure 12-2. CPU output clock timing. The timing of the CPU clock (Figure 12-2) depends on the selection of the CpuSel bit in the RegSysClock register and is given in Table 12-11. fmax is the frequency of fastest clock present in the circuit. Note that the tolerance on the 32 kHz depends on the selected clock source (see clock block documentation). CpuSel 0 1 f1 fmax/4 fmax f2 fmax 32 kHz Table 12-11. CPU clock timing parameters. Pins PB[5] and PB[4] can be used for S1 and S0 of the USRT (see USRT documentation) when the UsrtEnable bit is set in RegUsrtCtrl. The PB[5] and PB[4] then become open-drain. This overrides the values contained in PBOpen(5:4), PBOut(5:4) and PBDir(5:4). If there is no external pull-up resistor on these pins, internal pull-ups should be selected by setting PBPullup(5:4). When S0 is an output, the pin PB[4] takes the value of UsrtS0 in RegUsrtS0. When S1 is an output, the pin PB[5] takes the value of UsrtS1 in RegUsrtS1. Pins PB[6] and PB[7] can be used by the UART (see UART documentation). When UartEnTx in RegUartCtrl is set to 1, PB[6] is used as output signal Tx. When UartEnRx in RegUartCtrl is set to 1, PB[7] is used as input signal Rx. This overrides the values contained in PBOut(7:6) and PBDir(7:6). 12.7 Port B digital capabilities 12.7.1 Port B digital configuration The direction of each bit within Port B (input only or input/output) can be individually set using the RegPBDir register. If PBDir[x] = 1, both the input and output buffer are active on the corresponding Port B. If PBDir[x] = 0, the corresponding Port B pin is an input only and the output buffer is in high impedance. After reset (nresetpconf) Port B is in input only mode (PBDir[x] are reset to 0). The input values of Port B are available in RegPBIn (read only). Reading is always direct - there is no debounce function in Port B. In case of possible noise on input signals, a software debouncer with polling or an external hardware filter have to be realized. The input buffer is also active when the port is defined as output and the effective value on the pin can be read back. Data stored in RegPBOut are outputted at Port B if PBDir[x] is 1. The default values after reset is low (0). When a pin is in output mode (PBDir[x] is set to 1), the output can be a conventional CMOS (Push-Pull) or a Nchannel Open-drain, driving the output only low. By default, after reset (nresetpconf) the PBOpen[x] in RegPBOpen is cleared to 0 (push-pull). If PBOpen[x] in RegPBOpen is set to 1 then the internal P transistor in the output buffer is electrically removed and the output can only be driven low (PBOut[x]=0). When PBOut[x]=1, the pin is high Impedance. The internal pull-up or an external pull-up resistor can be used to drive to pin high. Note: Because the P transistor actually exists (this is not a real Open-drain output) the pull-up range is limited to VDD + 0.2V (avoid forward bias the P transistor / diode). Each bit can be set individually for pull-up or not using register RegPBPullup. Input is pulled up when its corresponding bit in this register is set to 1. Default status after (nresetpconf) is 1, which means with pull up. To © Semtech 2006 www.semtech.com 12-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver limit power consumption, pull-up resistors are only enabled when the associated pin is either a digital input or an Nchannel open-drain output with the pad set to 1. In the other cases (push-pull output or open-drain output driven low), the pull up resistors are disabled independent of the value in RegPBPullup. After power-on reset, the Port B is configured as an input port with pull-up. During power-on reset (see reset block documentation) however, the pin PB[1] is pulled down in stead of pulled up. Once the power-on reset completed, the pin PB[1] is pulled up, exactly as the other Port B pins. The input buffer is always active, except in analog mode. This means that the Port B input should be a valid digital value at all times unless the pin is set in analog mode. Violating this rule may lead to high power consumption. 12.7.2 sym VINH VINL VOH Port B digital function specification description Input high voltage Input low voltage Output high voltage min 0.7*VBAT VSS VBAT-0.4 VOL Output low voltage VSS RPU Cin Pull-up resistance Input capacitance 20 typ max VBAT 0.2*VBAT VBAT VSS+0.4 50 3.5 80 unit Comments V VBAT≥2.4V V VBAT≥2.4V V VBAT=1.2V, IOH =0.3mA VBAT=2.4V, IOH =5.0mA VBAT=4.5V, IOH =8.0mA V VBAT=1.2V, IOL =0.3mA VBAT=2.4V, IOL =12.0mA VBAT=4.5V, IOL =15.0mA kΩ pF Note 1 Note 1: this value is indicative only since it depends on the package. 12.8 Low power comparators If the low power comparator (CMPD) peripheral is present in the circuit, the signals on the pins PB[7:4] can be used as inputs for these low power comparators. Although the comparators are functional independent of the Port B configuration, it is recommended to set the pins that are used for the CMPD in analog mode without selecting any analog lines. This is to avoid high power consumption in the digital input buffer when analog or slowly varying digital signals are applied. © Semtech 2006 www.semtech.com 12-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 13. Port D 13.1 Features 13-2 13.2 Overview 13-2 13.3 Register map 13-2 13.4 Port D (PD) Operation 13-4 13.5 Port D electrical specification 13-5 © Semtech 2006 www.semtech.com 13-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 13.1 Features input / output port, 8 bits wide each bit can be set individually for input or output pull-ups are available in input mode snap-to-rail option in input mode 13.2 Overview Port D (PD) is a general purpose 8 bit input/output digital port. Figure 13-1 shows its structure. VBat 4 RegPDPullup[7:4] 4 logic RegPDPullup[3:0] 8 RegPDIn 8 RegPDOut 8 RegPDDir Vss Figure 13-1 : structure of PortD 13.3 Register map There are four registers in the Port D (PD), namely RegPDIn, RegPDOut, RegPDDir and RegPDPullup. Table 133 to Table 13-6 show the mapping of control bits and functionality of these registers. register name RegPDIn RegPDOut RegPDDir RegPDPullup Table 13-1 : PD registers © Semtech 2006 www.semtech.com 13-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Pos. 7:0 RegPDIn PDIn[7:0] Rw r Reset - Description pad PD[7:0] input value Table 13-2 : RegPDIn Pos. 7:0 RegPDOut PDOut[7:0] Rw rw Reset 0 nresetpconf Description pad PD[7:0] output value Table 13-3 : RegPDOut Pos. 7:0 RegPDDir PDDir[7:0] Rw rw Reset 0 nresetpconf Description pad PD[7:0] direction (0=input) Table 13-4 : RegPDDir Pos. 7 RegPDPullup PDSnapToRail[3] Rw rw Reset 1 nresetpconf 6 PDSnapToRail[2] rw 1 nresetpconf 5 PDSnapToRail[1] rw 1 nresetpconf 4 PDSnapToRail[0] rw 1 nresetpconf 3 PDPullup[3] rw 1 nresetpconf 2 PDPullup[2] rw 1 nresetpconf 1 PDPullup[1] rw 1 nresetpconf 0 PDPullup[0] rw 1 nresetpconf Description snap-to-rail for pad PD[7] and PD[6] (1=active) snap-to-rail for pad PD[5] and PD[4] (1=active) snap-to-rail for pad PD[3] and PD[2] (1=active) snap-to-rail for pad PD[1] and PD[0] (1=active) pullup for pad PD[7] and PD[6] (1=active) pullup for pad PD[5] and PD[4] (1=active) pullup for pad PD[3] and PD[2] (1=active) pullup for pad PD[1] and PD[0] (1=active) Table 13-5 : RegPDPullup © Semtech 2006 www.semtech.com 13-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 13.4 Port D (PD) Operation The direction of each pin of Port D (input or input/output) can be individually set by using the RegPDDir register. If PDDir[x] = 1, the output buffer on the corresponding Port D pin is enabled. After reset, Port D is in input only mode (PDDir[x] are reset to 0). The input buffer is always enabled independently from the RegPDDir contents. Output data: Data are stored in RegPDOut prior to output at Port D. Input data: The status of Port D is available in RegPDIn (read only). Reading is always direct - there is no digital debounce function associated with Port D. In case of possible noise on input signals, a software debouncer or an external filter must be realised. Pull-up/Snap to Rail: When configured as an input (PDDir[x]=0), pull-ups are available on every pin. The pull-up function of the pins is controlled two by two by the PDPullup and PDSnapToRail bits in the register RegPDPullup. When a bit PDPullup[x] is 0, the pull-ups on the pins PD[2x] and PD[2x+1] are disabled. When a bit PDPullup[x] is set to 1 and the bit PDSnapToRail[x] is set to 0, the pull-up resistor is connected to the pins PD[2x] and PD[2x+1]. When both PDPullup[x] and PDSnapToRail[x] are 1, the snap-to-rail function is active on the pins PD[2x] and PD[2x+1]. The snap-to-rail function connects a pullup or pulldown resistor to the input pin depending on the value forced on the input pin. This function can be used for instance when the input port is connected to a tristate bus. When the bus is floating, the pullup or pulldown maintains the bus in the last low impedance state before it became floating until another low impedance output is driving the bus. It also reduces the power consumption with respect to a classic pullup since it selects the pullup or pulldown resistor so that it confirms the detected input state. The function is summarised in the table below as a function of the different register settings. PDDir[2x(+1)] PDPullup[x] PDSnapToRail[x] 1 0 0 0 0 x 0 1 1 1 x x 0 1 1 (last) externally forced PD[2x(+1)] value x x x 0 1 PD[2x(+1)] pull resistor not connected not connected pullup pulldown pullup Table 13-6: Snap-to-rail and pullup function At power-on reset, Port D is configured as an input port with all pull-ups active. © Semtech 2006 www.semtech.com 13-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 13.5 Port D electrical specification sym VINH VINL VOH description Input high voltage Input low voltage Output high voltage min 0.7*VBAT VSS VBAT-0.4 VOL Output low voltage VSS RPU Cin Pull-up resistance Input capacitance 20 typ max VBAT 0.2*VBAT VBAT VSS+0.4 50 3.0 80 unit Comments V VBAT≥2.4V V VBAT≥2.4V V VBAT=1.2V, IOH =0.3mA VBAT=2.4V, IOH =5.0mA VBAT=4.5V, IOH =8.0mA V VBAT=1.2V, IOL =0.3mA VBAT=2.4V, IOL =12.0mA VBAT=4.5V, IOL =15.0mA kΩ pF Note 1 Note 1: this value is indicative only since it depends on the package. Table 13-7. Port D electrical specification © Semtech 2006 www.semtech.com 13-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 14. 14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8 14.8.1 14.8.2 14.8.3 14.8.4 14.9 Universal Asynchronous Receiver/Transmitter (UART) FEATURES .................................................................................................................................... 14-2 OVERVIEW .................................................................................................................................... 14-2 REGISTERS MAP............................................................................................................................ 14-2 INTERRUPTS MAP .......................................................................................................................... 14-3 UART BAUD RATE SELECTION ......................................................................................................... 14-3 UART ON THE RC OSCILLATOR OR EXTERNAL CLOCK SOURCE ........................................................... 14-3 UART ON THE CRYSTAL OSCILLATOR ............................................................................................... 14-4 FUNCTION DESCRIPTION ................................................................................................................ 14-5 Configuration bits........................................................................................................................ 14-5 Transmission .............................................................................................................................. 14-5 Reception ................................................................................................................................... 14-6 Interrupt or polling....................................................................................................................... 14-7 SOFTWARE HINTS.......................................................................................................................... 14-7 © Semtech 2006 www.semtech.com 14-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 14.1 Features • Full duplex operation with buffered receiver and transmitter. • Internal baudrate generator with 10 programmable baudrates (300 - 153600). • 7 or 8 bits word length. • Even, odd, or no-parity bit generation and detection • 1 stop bit • Error receive detection: Start, Parity, Frame and Overrun • Receiver echo mode • 2 interrupts (receive full and transmit empty) • Enable receive and/or transmit • Invert pad Rx and/or Tx 14.2 Overview The Uart pins are PB[7], which is used as Rx - receive and PB[6] as Tx - transmit. 14.3 Registers map register name RegUartCtrl RegUartCmd RegUartTx RegUartTxSta RegUartRx RegUartRxSta Table 14-1: Uart registers pos. 7 6 5-3 2 1 0 RegUartCmd SelXtal UartRcSel(2:0) UartPM UartPE UartWL rw r/w r r/w r/w r/w r/w reset 0 nresetglobal 0 000 nresetglobal 0 nresetglobal 0 nresetglobal 1 nresetglobal description Select input clock: 0 = RC/external, 1 = xtal Unused RC prescaler selection Select parity mode: 1 = odd, 0 = even Enable parity: 1 = with parity, 0 = no parity Select word length: 1 = 8 bits, 0 = 7 bits Table 14-2: RegUartCmd pos. 7 RegUartCtrl UartEcho rw r/w reset 0 nresetglobal 6 5 4 3 2-0 UartEnRx UartEnTx UartXRx UartXTx UartBR(2:0) r/w r/w r/w r/w r/w 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 101 nresetglobal description Enable echo mode: 1 = echo Rx->Tx, 0 = no echo Enable uart reception Enable uart transmission Invert pad Rx Invert pad Tx Select baud rate Table 14-3: RegUartCtrl © Semtech 2006 www.semtech.com 14-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7-0 RegUartTx UartTx rw r/w reset 00000000 nresetglobal description Data to be send Table 14-4: RegUartTx pos. 7-2 1 0 RegUartTxSta UartTxBusy UartTxFull rw r r r reset 000000 0 nresetglobal 0 nresetglobal description Unused Uart busy transmitting RegUartTx full Set by writing to RegUartTx Cleared when transferring RegUartTx into internal shift register Table 14-5: RegUartTxSta pos. 7-0 RegUartRx UartRx rw r reset 00000000 nresetglobal description Received data Table 14-6: RegUartRx pos. 7-6 5 4 3 2 RegUartRxSta UartRxSErr UartRxPErr UartRxFErr UartRxOErr rw r r r r r/c reset 00 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 1 0 UartRxBusy UartRxFull r r 0 nresetglobal 0 nresetglobal description Unused Start error Parity error Frame error Overrun error Cleared by writing RegUartRxSta Uart busy receiving RegUartRx full Cleared by reading RegUartRx Table 14-7: RegUartRxSta 14.4 Interrupts map interrupt source Irq_uart_Tx Irq_uart_Rx default mapping in the interrupt manager IrqHig(1) IrqHig(0) Table 14-8: Interrupts map 14.5 Uart baud rate selection In order to have correct baud rates, the Uart interface has to be fed with a stable and trimmed clock source. The clock source can be an external clock source, the RC oscillator or the crystal oscillator. The precision of the baud rate will depend on the precision of the selected clock source. 14.6 Uart on the RC oscillator or external clock source To select the external clock or RC oscillator for the Uart, the bit SelXtal in RegUartCmd has to be 0. The choice between the RC oscillator and the external clock source is made with the bit EnExtClock in RegSysClock. © Semtech 2006 www.semtech.com 14-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver In order to obtain a correct baud rate, the RC oscillator or external clock frequency have to be set to one of the frequencies given in the table below. The precision of the obtained baud rate is directly proportional to the frequency deviation of the used clock source with respect to the values in the table below. Frequency selection for correct Uart baud rates RC oscillator (Hz) External clock (Hz) 2’457’600 4’915’200 1’228’800 2’457’600 614’400 1’228’800 307’200 614’400 153’600 307’200 76’800 153’600 Table 14-9a For each of these frequencies, the baud rate can be selected with the bits UartBR(2:0) in RegUartCtrl and UartRcSel(2:0) in RegUartCmd as shown in Table 14-9. RC frequency (Hz) External clock freq. (Hz) UartRcSel UartBR 000 111 110 101 100 011 010 001 000 010 001 000 2457600 4915200 1228800 2457600 614400 1228800 307200 614400 153600 307200 76800 153600 153600 76800 38400 19200 9600 4800 2400 1200 600 300 76800 38400 19200 9600 4800 2400 1200 600 300 - 38400 19200 9600 4800 2400 1200 600 300 - 19200 9600 4800 2400 1200 600 300 - 9600 4800 2400 1200 600 300 - 4800 2400 1200 600 300 - Table 14-9: Uart baud rate with RC clock or external clock Note 1 : Although not documented here, the coding of the baud rate used in the circuits XE8801, XE8803 and XE8805 can also be used. Note 2 : The precision of the baud rate is directly proportional to the frequency deviation of the used clock from the ideal frequency given in the table. In order to increase the precision and stability of the RC oscillator, the DFLL (digital frequency locked loop) can be used with the crystal oscillator as a reference. 14.7 Uart on the crystal oscillator In order to use the crystal oscillator as the clock source for the Uart, the bit SelXtal in RegUartCmd has to be set. The crystal oscillator has to be enabled by setting the EnableXtal bit in RegSysClock. The baud rate selection is done using the UartBR bits as shown in Table 14-10. Xtal freq. (Hz) UartBR 011 010 001 000 32768 2400 1200 600 300 Table 14-10: Uart baud rate with Xtal clock © Semtech 2006 www.semtech.com 14-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Due to the odd ratio between the crystal oscillator frequency and the baud rate, the generated baud rate has a systematic error of –2.48%. 14.8 Function description 14.8.1 Configuration bits The configuration bits of the Uart serial interface can be found in the registers RegUartCmd and RegUartCtrl. The bit SelXtal is used to select the clock source (see chapter 14.5). The bits UartSelRc and UartBR select the baud rate (see chapter 14.5). The bits UartEnRx and UartEnTx are used to enable or disable the reception and transmission. The word length (7 or 8 data bits) can be chosen with UartWL. A parity bit is added during transmission or checked during reception if UartPE is set. The parity mode (odd or even) can be chosen with UartPM. Setting the bits UartXRx and UartXTx inverts the Rx respectively Tx signals. The bit UartEcho is used to send the received data automatically back. The transmission function becomes then: Tx = Rx XOR UartXRx XOR UartXTx. 14.8.2 Transmission In order to send data, the transmitter has to be enabled by setting the bit UartEnTx. Data to be sent have to be written to the register RegUartTx. The bit UartTxFull in RegUartTxSta then goes to 1, indicating to the transmitter that a new word is available. As soon as the transmitter has finished sending the previous word, it then loads the contents of the register RegUartTx to an internal shift register and clears the UartTxFull bit. An interrupt is generated on Irq_uart_Tx at the falling edge of the UartTxFull bit. The bit UartTxBusy in RegUartTxSta shows that the transmitter is busy transmitting a word. A timing diagram is shown in Figure 14-1. Data is sent LSB first. New data should be written to the register RegUartTx only while UartTxBusy is 0, otherwise data will be lost. © Semtech 2006 www.semtech.com 14-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Asynchronous Transmission write to RegUartTx word 1 RegUartTx reguarttx_shift word 1 shift clock Tx start b0 b1 b6/7 parity stop UartTxBusy UartTxFull Irq_uart_Tx Asynchronous Transmission (back to back) word 1 word 2 write to RegUartTx RegUartTx reguarttx_shift word 1 word 2 word 1 word 2 shift clock Tx start b0 b6/7 stop start UartTxBusy UartTxFull Irq_uart_Tx Figure 14-1. Uart transmission timing diagram. 14.8.3 Reception On detection of the start bit, the UartRxBusy bit is set. On detection of the stop bit, the received data are transferred from the internal shift register to the register RegUartRx. At the same time, the UartRxFull bit is set and an interrupt is generated on Irq_uart_Rx. This indicates that new data is available in RegUartRx. The timing diagram is shown in Figure 14-2. The UartRxFull bit is cleared when RegUartRx is read. If the register was not read before the receiver transfers a new word to it, the bit UartRxOErr (overflow error) is set and the previous contents of the register are lost. UartRxOErr is cleared by writing any data to RegUartRxSta. The bit UartRxSErr is set if a start error has been detected. The bit is updated at data transfer to RegUartRx. The bit UartRxPErr is set if a parity error has been detected, i.e. the received parity bit is not equal to the calculated parity of the received data. The bit is updated at data transfer to RegUartRx. The bit UartRxFErr in RegUartRxSta shows that a frame error has been detected. No stop bit has been detected. © Semtech 2006 www.semtech.com 14-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Asynchronous Reception read of RegUartRx (software) reguartrx_shift word 1 word 1 RegUartRx shift clock Rx start b0 b6/7 parity stop UartRxBusy UartRxFull Irq_uart_Rx Figure 14-2. Uart reception timing diagram. 14.8.4 Interrupt or polling The transmission and reception software can be driven by interruption or by polling the status bits. Interrupt driven reception: each time an Irq_uart_Rx interrupt is generated, a new word is available in RegUartRx. The register has to be read before a new word is received. Interrupt driven transmission: each time the contents of RegUartTx is transferred to the transmission shift register, an Irq_uart_Tx interrupt is generated. A new word can then be written to RegUartTx. Reception driven by polling: the UartRxFull bit is to be read and checked. When it is 1, the RegUartRx register contains new data and has to be read before a new word is received. Transmission driven by polling: the UartTxFull bit is to read and checked. When it is 0, the RegUartTx register is empty and a new word can be written to it. 14.9 Software hints Example of program for a transmission with polling: 1. The RegUartCmd register and the RegUartCtrl register are initialized (for example: 8 bit word length, odd parity, 9600 baud, enable Uart transmission). 2. Write a byte to RegUartTx. 3. Wait untill the UartTxFull bit in RegUartTxSta register equals 0. 4. Jump to 2 to writing the next byte if the message is not finished. 5. End of transmission. Example of program for a transmission with interrupt: 1. The RegUartCmd register and the RegUartCtrl register are initialized (for example: 8 bit word length, odd parity, 9600 baud, enable Uart transmission). 2. Write a byte to RegUartTx. 3. After an interrupt and if the message is not finished, jump to 2 4. End of transmission. © Semtech 2006 www.semtech.com 14-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Example of program for a reception with polling: 1. The RegUartCmd register and the RegUartCtrl register are initialized (for example: 8 bit word length, odd parity, 9600 baud, enable Uart reception). 2. Wait until the UartRxFull bit in the RegUartRxSta register equals 1. 3. Read the RegUartRxSta and check if there is no error. 4. Read data in RegUartRx. 5. If data is not equal to End-Of-Line, then jump to 2. 6. End of reception. Example of program for a reception with interrupt: 1. The RegUartCmd register and the RegUartCtrl register are initialized (for example: 8 bit word length, odd parity, 9600 baud, enable Uart reception). 2. When there is an interrupt, jump to 3 3. Read RegUartRxSta and check if there is no error. 4. Read data in RegUartRx. 5. If data is not equal to End-Of-Line, then jump to 2. 6. End of reception. © Semtech 2006 www.semtech.com 14-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 15. 15.1 15.2 15.3 15.4 15.5 15.6 15.7 15.8 Universal Synchronous Receiver/Transmitter (URST) FEATURES ........................................................................................................................................... 15-2 OVERVIEW ........................................................................................................................................... 15-2 REGISTER MAP ..................................................................................................................................... 15-2 INTERRUPTS MAP.................................................................................................................................. 15-4 CONDITIONAL EDGE DETECTION 1 .......................................................................................................... 15-4 CONDITIONAL EDGE DETECTION 2 .......................................................................................................... 15-4 INTERRUPTS OR POLLING ...................................................................................................................... 15-5 FUNCTION DESCRIPTION........................................................................................................................ 15-5 © Semtech 2006 www.semtech.com 15-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 15.1 Features The USRT implements a hardware support for software implemented serial protocols: Control of two external lines S0 and S1 (read/write). Conditional edge detection generates interrupts. S0 rising edge detection. S1 value is stored on S0 rising edge. S0 signal can be forced to 0 after a falling edge on S0 for clock stretching in the low state. S0 signal can be stretched in the low state after a falling edge on S0 and after a S1 conditional detection. 15.2 Overview The USRT block supports software universal synchronous receiver and transmitter mode interfaces. External lines S0 and S1 respectively correspond to clock line and data line. S0 is mapped to PB[4] and S1 to PB[5] when the USRT block is enabled. It is independent of RegPBdir (Port B can be input or output). When USRT is enabled, the configurations in port B for PB[4] and PB[5] are overwritten by the USRT configuration. Internal pull-ups can be used by setting the PBPullup[5:4] bits. Conditional edge detections are provided. RegUsrtS1 can be used to read the S1 data line from PB[5] in receive mode or to drive the output S1 line PB[5] by writing it when in transmit mode. It is advised to read S1 data when in receive mode from the RegUsrtBufferS1 register, which is the S1 value sampled on a rising edge of S0. 15.3 Register map Register name RegUsrtS1 RegUsrtS0 RegUsrtCtrl RegUsrtCond1 RegUsrtCond2 RegUsrtBufferS1 RegUsrtEdgeS0 Table 15-1: USRT registers Block configuration registers: pos. 7-1 0 RegUsrtS1 “0000000” UsrtS1 rw reset function r r/w 1 nresetglobal Unused Write: data S1 written to pad PB[5]), Read: value on PB[5] (not UsrtS1 value). Table 15-2: RegUsrtS1 © Semtech 2006 www.semtech.com 15-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7-1 0 RegUsrtS0 “0000000” UsrtS0 rw r r/w reset 1 nresetglobal function Unused Write: clock S0 written to pad PB[4], Read: value on PB[4] (not UsrtS0 value). Table 15-3: RegUsrtS0 The values that are read in the registers RegUsrtS1 and RegUsrtS0 are not necessarily the same as the values that were written in the register. The read value is read back on the circuit pins not in the registers themselves. Since the outputs are open drain, an external circuit on the circuit pins may force a value different from the register value. pos. 7-4 3 RegUsrtCtrl “0000” UsrtWaitS0 rw r r reset 0 nresetglobal 2 UsrtEnWaitCond1 r/w 0 nresetglobal 1 0 UsrtEnWaitS0 UsrtEnable r/w r/w 0 nresetglobal 0 nresetglobal function Unused Clock stretching flag (0=no stretching), cleared by writing RegUsrtBufferS1 Enable stretching on UsrtCond1 detection (0=disable) Enable stretching operation (0=disable) Enable USRT operation (0=disable) Table 15-4: RegUsrtCtrl pos. 7-1 0 RegUsrtCond1 “0000000” UsrtCond1 rw r r/c reset 0 nresetglobal function Unused State of condition 1 detection (1 =detected), cleared when written. Table 15-5: RegUsrtCond1 pos. 7-1 0 RegUsrtCond2 “0000000” UsrtCond2 rw r r/c reset 0 nresetglobal function Unused State of condition 2 detection (1 =detected), cleared when written. Table 15-6: RegUsrtCond2 pos. 7-1 0 RegUsrtBufferS1 “0000000” UsrtBufferS1 rw r r reset 0 nresetglobal function Unused Value on S1 at last S0 rising edge. Table 15-7: RegUsrtBufferS1 pos. 7-1 0 RegUsrtEdgeS0 “0000000” UsrtEdgeS0 rw r r reset 0 nresetglobal function Unused State of rising edge detection on S0 (1=detected). Cleared by reading RegUsrtBufferS1 Table 15-8: RegUsrtEdgeS0 © Semtech 2006 www.semtech.com 15-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 15.4 Interrupts map interrupt source Irq_cond2 Irq_cond1 default mapping in the interrupt manager RegIrqMid(7) RegIrqMid(6) Table 15-9: Interrupts map 15.5 Conditional edge detection 1 S1 S0 Figure 15-1: Condition 1 Condition 1 is satisfied when S0=1 at the falling edge of S1. The bit UsrtCond1 in RegUsrtCond1 is set when the condition 1 is detected and the USRT interface is enabled (UsrtEnable=1). Condition 1 is asserted for both modes (receiver and transmitter). The UsrtCond1 bit is read only and is cleared by all reset conditions and by writing any data to its address. Condition 1 occurrence also generates an interrupt on Irq_cond1. 15.6 Conditional edge detection 2 S1 S0 Figure 15-2: Condition 2 Condition 2 is satisfied when S0=1 at the rising edge of S1. The bit UsrtCond2 in RegUsrtCond2 is set when the condition 2 is detected and the USRT interface is enabled. Condition 2 is asserted for both modes (receiver and transmitter). The UsrtCond2 bit is read only and is cleared by all reset conditions and by writing any data to its address. Condition 2 occurrence also generates an interrupt on Irq_cond2. © Semtech 2006 www.semtech.com 15-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 15.7 Interrupts or polling In receive mode, there are two possibilities to detect condition 1 or 2: the detection of the condition can generate an interrupt or the registers can be polled (reading and checking the RegUsrtCond1 and RegUsrtCond2 registers for the status of USRT communication). 15.8 Function description The bit UsrtEnable in RegUsrtCtrl is used to enable the USRT interface and controls the PB[4] and PB[5] pins. This bit puts these two port B lines in the open drain configuration requested to use the USRT interface. If no external pull-ups are added on PB[4] and PB[5], the user can activate internal pull-ups by setting PBPullup[4] and PBPullup[5] in RegPBPullup. The bits UsrtEnWaitS0, UsrtEnWaitCond1, UsrtWaitS0 in RegUsrtCtrl are used for transmitter/receiver control of USRT interface. Figure 15-3 shows the unconditional clock stretching function which is enabled by setting UsrtEnWaitS0. S0 UsrtWaitS0 write Reg UsrtBufferS1 Figure 15-3: S0 Stretching (UsrtEnWaitS0=1) When UsrtEnWaitS0 is 1, the S0 line will be maintained at 0 after its falling edge (clock stretching). UsrtWaitS0 is then set to 1, indicating that the S0 line is forced low. One can release S0 by writing to the RegUsrtBufferS1 register. The same can be done in combination with condition 1 detection by setting the UsrtEnWaitCond1 bit. Figure 15-4 shows the conditional clock stretching function, which is enabled by setting UsrtEnWaitCond1. © Semtech 2006 www.semtech.com 15-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver S1 S0 UsrtWaitS0 write Reg UsrtBufferS1 Figure 15-4: Conditional stretching (UsrtEnWaitCond1=1) When UsrtEnWaitCond1 is 1, the S0 signal will be stretched in its low state after its falling edge if the condition 1 has been detected before (UsrtCond1=1). UsrtWaitS0 is then set to 1, indicating that the S0 line is forced low. One can release S0 by writing to the RegUsrtBufferS1 register. Figure 15-5 shows the sampling function implemented by the UsrtBufferS1 bit. The bit UsrtBufferS1 in RegUsrtBufferS1 is the value of S1 sampled on PB[4] at the last rising edge of S0. The bit UsrtEdgeS0 in RegUsrtEdgeS0 is set to one on the same S0 rising edge and is cleared by a read operation of the RegUsrtBufferS1 register. The bit therefor indicates that a new value is present in the RegUsrtBufferS1 which was not yet read. S1 S0 UsrtBufferS1 read Reg UsrtBufferS1 UsrtEdgeS0 Figure 15-5: S1 sampling © Semtech 2006 www.semtech.com 15-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 16 Serial Peripheral Interface 16.1 16.2 16.3 16.4 16.5 16.5.1 FEATURES .................................................................................................................................... 16-2 OVERVIEW ................................................................................................................................... 16-2 REGISTER MAP ............................................................................................................................. 16-2 INTERRUPTS MAP .......................................................................................................................... 16-4 FUNCTION DESCRIPTION ................................................................................................................ 16-4 SPI interface ............................................................................................................................... 16-4 16.5.1.1 16.5.1.2 16.5.1.3 General operation............................................................................................................................................. 16-4 Master/Slave synchronization .......................................................................................................................... 16-6 Software hints................................................................................................................................................... 16-7 16.5.2 General purpose port.................................................................................................................. 16-8 © Semtech 2006 www.semtech.com 16-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 16.1 Features The SPI block implements the following functionality’s: • Full duplex operating mode. • Master or slave configuration capability. • Separate transmissions data, shift data and receive data registers in order to perform back-to-back transmissions. • Four master mode frequencies available to generate serial clock. • Serial clock with programmable polarity and phase. • One enabled interrupt: SPI receive register full. • Overflow detection flag. • 4 I/O dedicated pads with 8mA drive and pull-up programmable. • Multi-slave configuration capability. • General purpose 4 bit wide digital input/output port mode. 16.2 Overview The SPI can communicate with other external SPI devices. It provides flexibility to communicate with different SPI compatible circuits from several manufacturers (Serial EEPROMs, display drivers, A/D converters, audio device). Four dedicated input or output pads are attached to the SPI block: MISO, MOSI, SCK, NSS. Six registers are used to run the SPI block. RegSpiControl, RegSpiStatus, RegSpiDataOut, RegSpiDataIn, RegSpiPullup, RegSpiDir are used to configure the communication settings, read the flags, write and read the exchanged data, and for the pad settings. The SPI device can also be used as a 4 bit general purpose input/output port. 16.3 Register map Register name RegSpiControl RegSpiStatus RegSpiDataOut RegSpiDataIn RegSpiPullup RegSpiDir Table 16-1: Address mapping for SPI When the peripheral is used as a general-purpose parallel I/O port, the SPI pads are mapped in the registers as follows (RegSpiDataOut, RegSpiDatain, RegSpiPullup and RegSpiDir): SPI pad names NSS MOSI MISO SCK SPI register bits 3 2 1 0 Table 16-2: Pin mapping in general purpose port mode © Semtech 2006 www.semtech.com 16-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Block configuration registers: pos. RegSpiControl Rw Reset Function Writing 1 clears transmission control counters. In master mode, this bit drives the NSS output pad. Unused in slave mode. It must be asserted (to 0) during a byte transfer. 0: SPI slave mode. 1: SPI master mode. peripheral configuration: 0: general purpose digital I/O port 1: SPI interface Controls the timing relationship between the serial clock and SPI data (cf. Figure 16-2 and Figure 16-3). Determines the idle-state of the SPI clock signal (cf. Figure 16-2 and Figure 16-3). Selects the baud rate in master mode. 00 => ckRcExt/2 01 => ckRcExt/8 10 => ckRcExt/16 11 => 4kHz 7 6 ClearCounter NotSlaveSelect w1 r/w 1 nresetglobal 5 SpiMaster r/w 1 nresetglobal 4 SpiEnable r/w 0 nresetglobal 3 ClockPhase r/w 1 nresetglobal 2 ClockPolarity r/w 0 nresetglobal 1-0 BaudRate r/w 00 nresetglobal Table 16-3: RegSpiControl Note that the precision of the 4kHz depends on the selected clock source (see documentation of the clock block). In slave mode, the fastest clock of the circuit should be at least 4 times faster than the baud rate of the master. pos. 7-3 2 RegSpiStatus -SpiOverflow Rw r r c1 Reset 00000 0 nresetglobal 1 SpiRxFull r 0 nresetglobal 0 SpiTxEmpty r w1 1 nresetglobal Function Unused This flag is set when a new byte is loaded in SpiDataIn before the previous byte was read. Writing 1 clears the flag. This flag is set each time a byte transfers from the shift register to SpiDataIn. It is cleared by reading SpiDataIn. This flag is cleared each time the CPU writes a byte in SpiDataOut. It is set when a byte transfers from SpiDataOut to the shift register. In the master mode, writing 1 to this bit performs the data transfer SpiDataOut to the shift register and enables the start of transmission. In the slave mode, the byte transfer is done automatically except for the very first byte after nresetglobal (cf.16.6 Software hints). Table 16-4: RegSpiStatus Pos. 7-0 RegSpiDataOut SpiDataOut[7:0] Rw r/w Reset 00000000 nresetglobal Function SPI mode: transmission data buffer I/O mode: output data (only bits 3:0, see Table 16-2) Table 16-5: RegSpiDataOut © Semtech 2006 www.semtech.com 16-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Pos. 7-0 RegSpiDataIn SpiDataIn[7:0] Rw r Reset 00000000 nresetglobal Function SPI mode: reception data byte I/O mode: input data (only bits 3:0, see Table 16-2, bits 7:4 read 0) Table 16-6: RegSpiDataIn Pos. 7-4 3-0 RegSpiPullup -SpiPullup Rw r r/w Reset 0000 1 nresetpconf Function Unused. Pullup configuration in both SPI and I/O mode (1=active). Mapping as in Table 16-2. Table 16-7: RegSpiPullup Note that pull-ups are disconnected independent from the value in RegSpiPullup if the corresponding pin is configured as an output. Pos. 7-4 3-0 RegSpiDir -SpiDir[3:0] Rw r r/w Reset 0000 0 nresetpconf Function Unused. I/O mode only (mapping in Table 16-2) 1: output enabled 0: output disabled Table 16-8: RegSpiDir 16.4 Interrupts map interrupt source Irq_spi(0) Default mapping in the interrupt manager irq_bus(21) Table 16-9: Interrupts map Irq_spi(0) interrupt is activated at the assertion of the SpiRxFull flag (cf. RegSpiStatus description). 16.5 Function description Depending on the value of the bit SpiEnable in the RegSpiControl register, the SPI peripheral can work either as a real SPI interface (master or slave) or as a general purpose 4 bit wide digital input/output port. 16.5.1 SPI interface 16.5.1.1 General operation The peripheral is configured as an SPI by setting the bit SpiEnable=1 in RegSpiControl. The bit SpiMaster in RegSpiControl selects the master or slave mode. The SPI interface supports 4 physical wires between one master device and one or more slave devices (Figure 16-1): MISO (Master In Slave Out), MOSI (Master Out Slave In), SCK (Serial Clock), NSS (Slave Select). A byte transmission performs a rotate operation between the value stored in the 8 bit shift register of the master device and the value stored in the 8 bit shift register of the (selected) slave device. The SCK line is used to synchronize both SPI interfaces. © Semtech 2006 www.semtech.com 16-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver The transmission format depends on the two configuration bits ClockPhase and ClockPolarity. These 2 bits should be configured in the same way in the slave and the master device in order to properly run the SPI transmission. The transmission baud rate depends on the two bits BaudRate, these 2 bits are only used in the master device to generate the serial clock SCK signal at the selected frequency. Data are transferred in a duplex way from master to slave through the MOSI wire and from slave to master through the MISO wire. Data are always sent most significant bit first. Each SPI device sequentially operates in two times: one clock edge to sample the received bit, and the other clock edge to shift the byte inside the shift register. The NSS signal is software controlled. It must be driven by the SPI master device by writing to the bit NotSlaveSelect in RegSpiControl. NSS should remain low during the byte transmission. Shift register MISO MISO MOSI MOSI SCK SCK NSS NSS Shift register Baud Rate Generator Slave Master Figure 16-1: Connection of master and slave device The next figure shows the timing diagrams for a SPI transmission with ClockPhase equal to 0. This means the active state of the serial clock SCK signal occurs on the 2nd half of the SCK cycle. 1 2 3 4 5 6 7 8 MOSI MSB bit 6 bit 5 bit 4 bit 3 bit 2 bit 1 LSB MISO MSB bit 6 bit 5 bit 4 bit 3 bit 2 bit 1 LSB SCK cycle SCK (CPOL=0) SCK (CPOL=1) Shift Samp le Shift Samp le Shift Samp le Shift Samp le Shift Samp le Shift Samp le Shift Samp le Shift Samp le NSS Figure 16-2: SPI transmission format with ClockPhase=0 The next figure shows the timing diagrams for a SPI transmission with ClockPhase equal to 1. This means the st active state of the serial clock SCK signal occurs on the 1 half of the SCK cycle. © Semtech 2006 www.semtech.com 16-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 1 2 3 4 5 6 7 8 MOSI MSB bit 6 bit 5 bit 4 bit 3 bit 2 bit 1 LSB MISO MSB bit 6 bit 5 bit 4 bit 3 bit 2 bit 1 LSB SCK cycle SCK (CPOL=0) SCK (CPOL=1) Shift Samp le Shift Samp le Shift Samp le Shift Samp le Shift Samp le Shift Samp le Shift Samp le Shift Samp le NSS Figure 16-3: SPI transmission format with ClockPhase=1 Note: for both cases, it is not required to toggle the NSS signal back to high and back to low between each byte transmitted. The SPI interface can be operated by polling the RegSpiStatus register or interrupt driven. The interrupt is active on reception of a new byte. In case of a multi-slave configuration, any digital output pin of any parallel port can be used to select the different slaves. In some cases, it might be easier to have DC signals on these pins and to derive the timing from the single NSS pin independently from the selected slave. This can be realized by combining the NSS wire from the master device with signals coming from an output port as shown in Figure 16-4. Pull-up resistors can be added on the input pads (MISO in master mode, MOSI, NSS and SCK in slave mode) by setting the corresponding bits in RegSpiPullup. Use Table 16-2 for the correspondence between the pads and register bits. 16.5.1.2 Master/Slave synchronization In the master mode, a transmission is started by writing a 1 to the bit SpiTxEmpty. This automatically loads the data of the register RegSpiDataOut to the shift register and starts the clock and shifting. At the end of the transmission, the clock stops and the received data are copied to RegSpiDataIn. The bit SpiTxEmpty should not be asserted while the previous transmission is still running, otherwise, the transmitted and received data will be corrupted. In slave mode, the fastest clock in the circuit should be at least 4 times faster than the baud rate of the transmission. The transmission is synchronized by the NSS input signal. While the NSS signal is high, the counters controlling the transmission are reset. The reception starts at the first clock cycle after the falling edge of NSS. At the end of the transmission, the received data are copied to RegSpiDataIn and the contents of the register RegSpiDataOut are copied automatically to the shift register. The data in the shift register can be overwritten by writing 1 to SpiTxEmpty. This should not be done while a transmission is running, otherwise, the transmitted and received data will be corrupted. The counters controlling the timing of the transmission can be reset by writing a 1 to the ClearCounters bit in the RegSpiControl register. In master mode, it restarts a complete transmission cycle. In slave mode, it has the same effect as a rising edge of NSS. This bit should be used with caution. © Semtech 2006 www.semtech.com 16-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Shift register MISO MISO MOSI MOSI SCK SCK NSS Shift register NSS Baud Rate Generator Slave 1 digital output pad 1 MISO digital output pad 2 Shift register MOSI digital output pad 3 SCK NSS Master Slave 2 MISO Shift register MOSI SCK NSS Slave N Figure 16-4: Connection of master and slaves in the case of a multi slave configuration 16.6 Software hints 16.6.1 Master mode The following routine must be executed by the CoolRISC in order to run properly the SPI interface in master mode: INITIALIZATION 1- Write RegSpiControl to enable the SPI interface and to configure the master mode (configured by default) and communication settings. 2- Write the data to send in the RegSpiDataOut register. The SpiTxEmpty flag toggles low. 3- Write 1 to the SpiTxEmpty bit to load the shift register with the value inside the RegSpiDataOut register and to start the byte transmission. The SpiTxEmpty flag toggles back high. 4- Write the next data to send to the RegSpiDataOut register. The SpiTxEmpty flag toggles low. © Semtech 2006 www.semtech.com 16-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver NORMAL OPERATION 5- Wait for the end of transmission, i.e. SpiRxFull flag is one or the interrupt of the receiver is asserted. 6- Write 1 to the SpiTxEmpty bit to load the shift register with the value inside the RegSpiDataOut register and to start the byte transmission. The SPI SpiTxEmpty flag toggles back high. 7- Read RegSpiDataIn. The SpiRxFull flag returns to 0. 8- Write the next data to send in the RegSpiDataOut register. The SpiTxEmpty flag toggles low. 9- Jump to 5. 16.6.2 Slave mode The following routine must be executed by the CoolRISC in order to run properly the SPI interface in slave mode: INITIALIZATION 1- Write RegSpiControl to enable the SPI interface and to configure the slave mode and communication settings (as configured in the master device). 2- Write the data to send in the RegSpiDataOut register. The SpiTxEmpty flag toggles low. 3- Write 1 to SpiTxEmpty bit to load the shift register with the value of the RegSpiDataOut register. The SpiTxEmpty flag toggles back high. This load is only required for the first byte to transmit after nresetglobal. It is automatic for the following bytes. 4- Write the next data to send in the RegSpiDataOut register. The SpiTxEmpty flag toggles low. NORMAL OPERATION 5- Wait for the end of transmission, i.e. the SpiRxFull flag is one or the interrupt of the receiver asserted. The shift register is automatically loaded with the value inside the RegSpiDataOut register and the SpiTxEmpty flag toggles back high. 6- Read RegSpiDataIn. The SpiRxFull flag returns to 0. 7- Write the next data to send in the RegSpiDataOut register. The SpiTxEmpty flag toggles low. 8- Jump to 5. 16.6.3 General purpose port. This mode is enabled when SpiEnable value is 0 (default value). The SPI dedicated pads are used as a general purpose 4 bit input/output digital port. Next figure shows the structure of the SPI in this mode. © Semtech 2006 www.semtech.com 16-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver vdd SPI spi_pullup[3:0] spi_out[3:0] RegSpiPullup RegSpiDataOut spi_dir[3:0] RegSpiDir spi_in[3:0] RegSpiDataIn Figure 16-5: Structure of the SPI general purpose port The direction of each bit within the SPI (input or input/output) can be individually set by using the RegSpiDir register. If RegSpiDir[x] = 1, the corresponding SPI pad becomes an output. After reset (nresetpconf), the SPI pads are in input configuration (RegSpiDir[x] are reset to 0). In output configuration, the data are stored in RegSpiDataOut prior to output at SPI pads. In input configuration, the status of SPI pads is available in RegSpiIn (read only). Reading is always direct –there is no digital debounce function associated with SPI pads. In case of possible noise on input signals, a software debouncer or an external filter must be realized. If a bit in RegSpiPullup is set, the pull-up of the corresponding input pad is active. The pull-ups are disabled in output configuration independently of the RegSpiPullup content. By default after reset, the SPI pads are configured as input ports with all pull-ups active. Note that the pull-up resistors can be used for the input pads in SPI mode also. Next table shows the link between SPI pads and SPI data bits. SPI pad names NSS MOSI MISO SCK SPI data bits SpiDataOut[3] or SpiDataIn[3] SpiDataOut[2] or SpiDataIn[2] SpiDataOut[1] or SpiDataIn[1] SpiDataOut[0] or SpiDataIn[0] Table 16-10: SPI pad/bit relationship © Semtech 2006 www.semtech.com 16-9 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 17. Acquisition 17.1 17.2 17.3 17.4 17.4.1 17.4.2 17.4.3 17.5 17.6 17.6.1 17.6.2 17.6.3 17.6.4 17.7 17.7.1 17.7.2 17.7.3 17.7.4 17.7.5 17.7.6 17.7.7 17.7.8 17.8 17.8.1 17.8.2 17.8.3 Chain ZOOMINGADC™ FEATURES ....................................................................................................... 17-2 OVERVIEW ................................................................................................................................. 17-2 REGISTER MAP ........................................................................................................................... 17-2 ZOOMINGADC™ DESCRIPTION ................................................................................................... 17-4 Acquisition Chain ..................................................................................................................... 17-4 Peripheral Registers................................................................................................................. 17-5 Continuous-Time vs. On-Request ............................................................................................ 17-7 INPUT MULTIPLEXERS ................................................................................................................. 17-7 PROGRAMMABLE GAIN AMPLIFIERS.............................................................................................. 17-8 PGA & ADC Enabling............................................................................................................. 17-10 PGA1 ..................................................................................................................................... 17-10 PGA2 ..................................................................................................................................... 17-10 PGA3 ..................................................................................................................................... 17-10 ADC CHARACTERISTICS ........................................................................................................... 17-11 Conversion Sequence............................................................................................................ 17-11 Sampling Frequency .............................................................................................................. 17-12 Over-Sampling Ratio.............................................................................................................. 17-12 Elementary Conversions ........................................................................................................ 17-13 Resolution .............................................................................................................................. 17-13 Conversion Time & Throughput ............................................................................................. 17-14 Output Code Format .............................................................................................................. 17-14 Power Saving Modes ............................................................................................................. 17-16 SPECIFICATIONS AND MEASURED CURVES ................................................................................. 17-16 Default Settings...................................................................................................................... 17-16 Specifications......................................................................................................................... 17-17 Linearity ................................................................................................................................. 17-18 17.8.3.1 Integral non-linearity.................................................................................................................................... 17-18 17.8.3.2 Differential non-linearity .............................................................................................................................. 17-22 17.8.4 17.8.5 17.8.6 17.8.7 17.9 17.9.1 17.9.2 17.9.3 17.9.4 17.9.5 Noise...................................................................................................................................... 17-23 Gain Error and Offset Error .................................................................................................... 17-24 Power Consumption............................................................................................................... 17-25 Power Supply Rejection Ratio................................................................................................ 17-27 APPLICATION HINTS .................................................................................................................. 17-28 Input Impedance .................................................................................................................... 17-28 PGA Settling or Input Channel Modifications.......................................................................... 17-28 PGA Gain & Offset, Linearity and Noise................................................................................. 17-28 Frequency Response ............................................................................................................. 17-29 Power Reduction.................................................................................................................... 17-29 © Semtech 2006 www.semtech.com 17-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 17.1 ZoomingADC™ Features The ZoomingADC™ is a complete and versatile low-power analog front-end interface typically intended for sensing applications. The key features of the ZoomingADC™ are: Programmable 6 to 16-bit dynamic range oversampled ADC • Flexible gain programming between 0.5 and 1000 • Flexible and large range offset compensation • 4-channel differential or 8-channel single-ended input multiplexer • 2-channel differential reference inputs • Power saving modes • Direct interfacing to CoolRisc™ microcontroller 17.2 Overview Analog Inputs MUX fS 0 1 2 3 4 5 6 7 VIN fS PGA1 PGA2 VD1 GD1 Input Selection MUX 0 Reference 1 Inputs 2 PGA3 VIN,ADC VD2 GD2 GD3 OFF2 OFF3 Offset2 Offset3 ADC 16 VREF 3 Reference Selection Gain1 Gain2 Gain3 ZOOM Figure 17-1. ZoomingADC™ general functional block diagram The total acquisition chain consists of an input multiplexer, 3 programmable gain amplifier stages and an oversampled A/D converter. The reference voltage can be selected on two different channels. Two offset compensation amplifiers allow for a wide offset compensation range. The programmable gain and offset give the ability to zoom in on a small portion of the reference voltage defined input range. 17.3 Register map There are eight registers in the acquisition chain (AC), namely RegAcOutLsb, RegAcOutMsb, RegAcCfg0, RegAcCfg1, RegAcCfg2, RegAcCfg3, RegAcCfg4 and RegAcCfg5. Table 17-2 to Table 17-9 show the mapping of control bits and functionality of these registers while Table 17-1 gives an overview of these eight. The register map only gives a short description of the different configuration bits. More detailed information is found in subsequent sections. © Semtech 2006 www.semtech.com 17-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver register name RegAcOutLsb RegAcOutMsb RegAcCfg0 RegAcCfg1 RegAcCfg2 RegAcCfg3 RegAcCfg4 RegAcCfg5 Table 17-1: AC registers pos. 7:0 RegAcOutLsb Out[7:0] rw r reset 00000000 nresetglobal description LSB of the output code Table 17-2: RegAcOutLsb pos. 7:0 RegAcOutMsb Out[15:8] rw r reset 00000000 nresetglobal description MSB of the output code Table 17-3: RegAcOutMsb pos. 7 6:5 RegAcCfg0 Start SET_NELCONV[1:0] rw w r0 rw reset 0 nresetglobal 01 nresetglobal 4:2 SET_OSR[2:0] rw 010 nresetglobal 1 0 CONT reserved rw rw 0 nresetglobal 0 nresetglobal description starts a conversion sets the number of elementary conversions sets the oversampling rate of an elementary conversion continuous conversion mode Table 17-4: RegAcCfg0 pos. 7:6 5:4 3:0 RegAcCfg1 IB_AMP_ADC[1:0] IB_AMP_PGA[1:0] ENABLE[3:0] rw rw rw rw reset 11 nresetglobal 11 nresetglobal 0000 nresetblobal description Bias current selection of the ADC converter Bias current selection of the PGA stages Enables the different PGA stages and the ADC Table 17-5: RegAcCfg1 pos. 7:6 5:4 3:0 RegAcCfg2 FIN[1:0] PGA2_GAIN[1:0] PGA2_OFFSET[3:0] rw rw rw rw reset 00 nresetglobal 00 nresetglobal 0000 nresetglobal description Sampling frequency selection PGA2 stage gain selection PGA2 stage offset selection Table 17-6: RegAcCfg2 pos. 7 6:0 RegAcCfg3 PGA1_GAIN PGA3_GAIN[6:0] rw rw rw reset 0 nresetglobal 0000000 nresetglobal description PGA1 stage gain selection PGA3 stage gain selection Table 17-7: RegAcCfg3 © Semtech 2006 www.semtech.com 17-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7 6:0 RegAcCfg4 reserved PGA3_OFFSET[6:0] rw r rw reset 0 0000000 nresetglobal description Unused PGA3 stage offset selection Table 17-8: RegAcCfg4 pos. 7 6 5:1 0 RegAcCfg5 BUSY DEF AMUX[4:0] VMUX rw r w r0 rw rw reset 0 nresetglobal 0 00000 nresetglobal 0 nresetglobal description Activity flag Selects default configuration Input channel configuration selector Reference channel selector Table 17-9: RegAcCfg5 17.4 ZoomingADC™Description Figure 17-2 gives a more detailed description of the acquisition chain. 17.4.1 Acquisition Chain Figure 17-1 shows the general block diagram of the acquisition chain (AC). A control block (not shown in Figure 17-1) manages all communications with the CoolRisc™ microcontroller. Analog inputs can be selected among eight input channels, while reference input is selected between two differential channels. The core of the zooming section is made of three differential programmable amplifiers (PGA). After selection of a combination of input and reference signals VIN and VREF, the input voltage is modulated and amplified through stages 1 to 3. Fine gain programming up to 1'000V/V is possible. In addition, the last two stages provide programmable offset. Each amplifier can be bypassed if needed. The output of the PGA stages is directly fed to the analog-to-digital converter (ADC), which converts the signal VIN,ADC into digital. Like most ADCs intended for instrumentation or sensing applications, the ZoomingADC™ is an over-sampled converter (See Note1). The ADC is a so-called incremental converter, with bipolar operation (the ADC accepts both positive and negative input voltages). In first approximation, the ADC output result relative to full-scale (FS) delivers the quantity: OUT ADC V IN , ADC ≅ FS / 2 V REF / 2 (Eq. 1) in two's complement (see Sections 17.4 and 17.7 for details). The output code OUTADC is -FS/2 to +FS/2 for VIN,ADC ≅ -VREF/2 to +VREF/2 respectively. As will be shown in section 17.6, VIN,ADC is related to input voltage VIN by the relationship: V IN , ADC = GDTOT ⋅ V IN − GDoff TOT ⋅ V REF (V) (Eq. 2) where GDTOT is the total PGA gain, and GDoffTOT is the total PGA offset. 1 Note: Over-sampled converters are operated with a sampling frequency f much higher than the input signal's Nyquist rate (typically f is 20S S 1'000 times the input signal bandwidth). The sampling frequency to throughput ratio is large (typically 10-500). These converters include digital decimation filtering. They are mainly used for high resolution, and/or low-to-medium speed applications. © Semtech 2006 www.semtech.com 17-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 0 1 2 3 AC_A 4 5 6 7 fS MUX Inputs PGA2 PGA3 GD1 GD2 GD3 OFF2 OFF3 VIN,ADC VIN ADC VREF MUX 0 1 AC_R 2 fS PGA1 3 5 2 4 7 Acquisition Chain 7 Register Bank RegACCfg5 RegACCfg4 8 RegACCfg3 RegACOutLSB RegACCfg2 RegACOutMSB RegACCfg1 RegACCfg0 Power Saving Modes PGA Enabling 8 ADC Busy Flag Default Settings Conversion Start Nbr of Elementary Cycles Over-Sampling Ratio Continuous vs. On-Request Sampling Frequency fS Figure 17-2. ZoomingADC™ detailed functional block diagram 17.4.2 Peripheral Registers Figure 17-2 shows a detailed functional diagram of the ZoomingADC™. In Table 17-10 the configuration of the peripheral registers is detailed. The system has a bank of eight 8-bit registers: six registers are used to configure the acquisition chain (RegAcCfg0 to 5), and two registers are used to store the output code of the analog-to-digital conversion (RegAcOutMsb & Lsb). The register coding of the ADC parameters and performance characteristics are detailed in Section 17.7. Table 17-10. Peripheral registers to configure the acquisition chain (AC) and to store the analog-to-digital conversion (ADC) result Register Name Bit Position 7 6 5 4 RegAcOutLsb 2 1 0 OUT[7:0] RegAcOutMsb RegAcCfg0 Default values: RegAcCfg1 Default values: 3 OUT[15:8] START 0 SET_NELC[1:0] 01 SET_OSR[2:0] 010 IB_AMP_ADC[1:0] IB_AMP_PGA[1:0] 11 11 © Semtech 2006 CONT 0 ENABLE[3:0] 0001 TEST 0 www.semtech.com 17-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Cont’ Register Name RegAcCfg2 Default values: RegAcCfg3 Default values: RegAcCfg4 Default values: RegAcCfg5 Default values: Bit Position 7 6 FIN[1:0] 00 PGA1_G 0 0 BUSY 0 DEF 0 5 4 3 2 1 PGA2_GAIN[1:0] PGA2_OFFSET[3:0] 00 0000 PGA3_GAIN[6:0] 0000000 PGA3_OFFSET[6:0] 0000000 AMUX[4:0] 00000 0 VMUX 0 With: • OUT: (r) digital output code of the analog-to-digital converter. (MSB = OUT[15]) • START: (w) setting this bit triggers a single conversion (after the current one is finished). This bit always reads back 0. SET_NELC[1:0] • SET_NELC: (rw) sets the number of elementary conversions to 2 . To compensate for offsets, the input signal is chopped between elementary conversions (1,2,4,8). (3+SET_OSR[2:0]) . OSR = 8, 16, 32, ..., • SET_OSR: (rw) sets the over-sampling rate (OSR) of an elementary conversion to 2 512, 1024. • CONT: (rw) setting this bit starts a conversion. A new conversion will automatically begin as long as the bit remains at 1. • TEST: bit only used for test purposes. In normal mode, this bit is forced to 0 and cannot be overwritten. • IB_AMP_ADC: (rw) sets the bias current in the ADC to 0.25*(1+ IB_AMP_ADC[1:0]) of the normal operation current (25, 50, 75 or 100% of nominal current). To be used for low-power, low-speed operation. • IB_AMP_PGA: (rw) sets the bias current in the PGAs to 0.25*(1+IB_AMP_PGA[1:0]) of the normal operation current (25, 50, 75 or 100% of nominal current). To be used for low-power, low-speed operation. • ENABLE: (rw) enables the ADC modulator (bit 0) and the different stages of the PGAs (PGAi by bit i=1,2,3). PGA stages that are disabled are bypassed. • FIN: (rw) These bits set the sampling frequency of the acquisition chain. Expressed as a fraction of the oscillator frequency, the sampling frequency is given as: 00 Æ 1/4 fRC, 01 Æ 1/8 fRC, 10 Æ 1/32 fRC, 11Æ ~8kHz. • PGA1_GAIN: (rw) sets the gain of the first stage: 0 Æ 1, 1 Æ 10. • PGA2_GAIN: (rw) sets the gain of the second stage: 00 Æ 1, 01 Æ 2, 10 Æ 5, 11 Æ 10. • PGA3_GAIN: (rw) sets the gain of the third stage to PGA3_GAIN[6:0]⋅1/12. • PGA2_OFFSET: (rw) sets the offset of the second stage between –1 and +1, with increments of 0.2. The MSB gives the sign (0 → positive, 1 → negative); amplitude is coded with the bits PGA2_OFFSET[5:0]. • PGA3_OFFSET: (rw) sets the offset of the third stage between –5.25 and +5.25, with increments of 1/12. The MSB gives the sign (0 → positive, 1 → negative); amplitude is coded with the bits PGA3_OFFSET[5:0]. • BUSY: (r) set to 1 if a conversion is running. • DEF: (w) sets all values to their defaults (PGA disabled, max speed, nominal modulator bias current, 2 elementary conversions, over-sampling rate of 32) and starts a new conversion without waiting the end of the preceding one. • AMUX(4:0): (rw) AMUX[4] sets the mode (0 Æ 4 differential inputs, 1 Æ 7 inputs with A(0) = common reference) AMUX(3) sets the sign (0 Æ straight, 1Æ cross) AMUX[2:0] sets the channel. • VMUX: (rw) sets the differential reference channel (0 Æ R(1) and R(0), 1 Æ R(3) and R(2)). (r = read; w = write; rw = read & write) © Semtech 2006 www.semtech.com 17-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 17.4.3 Continuous-Time vs. On-Request The ADC can be operated in two distinct modes: "continuous-time" and "on-request" modes (selected using the bit CONT). In "continuous-time" mode, the input signal is repeatedly converted into digital. After a conversion is finished, a new one is automatically initiated. The new value is then written in the result register, and the corresponding internal trigger pulse is generated. This operation is sketched in Figure 17-3. The conversion time in this case is defined as TCONV. TCONV Internal Trig Ouput Code RegACOut[15:0] BUSY IRQ Figure 17-3. ADC "continuous-time" operation TCONV Internal Trig Request START Ouput Code RegACOut[15:0] BUSY IRQ Figure 17-4. ADC "on-request" operation In the "on-request" mode, the internal behavior of the converter is the same as in the "continuous-time" mode, but the conversion is initiated on user request (with the START bit). As shown in Figure 17-4, the conversion time is also TCONV. 17.5 Input Multiplexers The ZoomingADC™ has eight analog inputs AC_A(0) to AC_A(7) and four reference inputs AC_R(0) to AC_R(3). Let us first define the differential input voltage VIN and reference voltage VREF respectively as: VIN = VINP − VINN (V) (Eq. 3) VREF = VREFP − VREFN (V) (Eq. 4) and: As shown in Table 17-11 the inputs can be configured in two ways: either as 4 differential channels (VIN1 = AC_A(1) - AC_A(0),..., VIN4 = AC_A(7) - AC_A(6)), or AC_A(0) can be used as a common reference, providing 7 signal paths all referenced to AC_A(0). The control word for the analog input selection is AMUX[4:0]. Notice that the bit AMUX[3] controls the sign of the input voltage. © Semtech 2006 www.semtech.com 17-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver AMUX[4:0] (RegAcCfg5[5:1]) VINP VINN AMUX[4:0] (RegAcCfg5[5:1]) VINP VINN 00x00 00x01 00x10 00x11 AC_A(1) AC_A(3) AC_A(5) AC_A(7) AC_A(0) AC_A(2) AC_A(4) AC_A(6) 01x00 01x01 01x10 01x11 AC_A(0) AC_A(2) AC_A(4) AC_A(6) AC_A(1) AC_A(3) AC_A(5) AC_A(7) 10000 10001 10010 10011 10100 10101 10110 10111 AC_A(0) AC_A(1) AC_A(2) AC_A(3) AC_A(4) AC_A(5) AC_A(6) AC_A(7) AC_A(0) 11000 11001 11010 11011 11100 11101 11110 11111 AC_A(0) AC_A(0) AC_A(1) AC_A(2) AC_A(3) AC_A(4) AC_A(5) AC_A(6) AC_A(7) Table 17-11. Analog input selection Similarly, the reference voltage is chosen among two differential channels (VREF1 = AC_R(1)-AC_R(0) or VREF2 = AC_R(3)-AC_R(2)) as shown in Table 17-12. The selection bit is VMUX. The reference inputs VREFP and VREFN (common-mode) can be up to the power supply range. VMUX (RegAcCfg5[0]) 0 1 VREFP VREFN AC_R(1) AC_R(3) AC_R(0) AC_R(2) Table 17-12. Analog Reference input selection 17.6 Programmable Gain Amplifiers As seen in Figure 17-1, the zooming function is implemented with three programmable gain amplifiers (PGA). These are: • PGA1: coarse gain tuning • PGA2: medium gain and offset tuning • PGA3: fine gain and offset tuning All gain and offset settings are realized with ratios of capacitors. The user has control over each PGA activation and gain, as well as the offset of stages 2 and 3. These functions are examined hereafter. ENABLE[3:0] Block xxx0 xxx1 xx0x xx1x x0xx x1xx 0xxx 1xxx ADC disabled ADC enabled PGA1 disabled PGA1 enabled PGA2 disabled PGA2 enabled PGA3 disabled PGA3 enabled Table 17-13. ADC & PGA enabling © Semtech 2006 www.semtech.com 17-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver PGA1 Gain GD1 (V/V) 1 10 PGA1_GAIN 0 1 Table 17-14. PGA1 Gain Settings PGA2 Gain GD2 (V/V) 1 2 5 10 PGA2_GAIN[1:0] 00 01 10 11 Table 17-15. PGA2 gain settings PGA2_OFFSET[3:0] 0000 0001 0010 0011 0100 0101 1001 1010 1011 1100 1101 PGA2 Offset GDoff2 (V/V) 0 +0.2 +0.4 +0.6 +0.8 +1 -0.2 -0.4 -0.6 -0.8 -1 Table 17-16. PGA2 offset settings PGA3 Gain GD3 (V/V) 0 1/12(=0.083) ... 6/12 ... 12/12 16/12 PGA3_GAIN[6:0] 0000000 0000001 ... 0000110 ... 0001100 0010000 ... 0100000 ... 1000000 ... 1111111 32/12 64/12 127/12(=10.58) Table 17-17. PGA3 gain settings © Semtech 2006 www.semtech.com 17-9 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver PGA3_OFFSET[6:0] 0000000 0000001 0000010 ... 0010000 ... 0100000 ... 0111111 1000000 1000001 1000010 ... 1010000 ... 1100000 ... 1111111 PGA3 Offset GDoff3 (V/V) 0 +1/12(=+0.083) +2/12 ... +16/12 ... +32/12 ... +63/12(=+5.25) 0 -1/12(=-0.083) -2/12 ... -16/12 ... -32/12 ... -63/12(=-5.25) Table 17-18. PGA3 offset settings 17.6.1 PGA & ADC Enabling Depending on the application objectives, the user may enable or bypass each PGA stage. This is done using the word ENABLE and the coding given in Table 17-13. To reduce power dissipation, the ADC can also be inactivated while idle. 17.6.2 PGA1 The first stage can have a buffer function (unity gain) or provide a gain of 10 (see Table 17-14). The voltage VD1 at the output of PGA1 is: V D1 = GD1 ⋅ V IN (V) (Eq. 5) where GD1 is the gain of PGA1 (in V/V) controlled with the bit PGA1_GAIN. 17.6.3 PGA2 The second PGA has a finer gain and offset tuning capability, as shown in Table 17-15 and Table 17-16. The voltage VD2 at the output of PGA2 is given by: V D 2 = GD2 ⋅ V D1 − GDoff 2 ⋅ V REF (V) (Eq. 6) where GD2 and GDoff2 are respectively the gain and offset of PGA2 (in V/V). These are controlled with the words PGA2_GAIN[1:0] and PGA2_OFFSET[3:0] . 17.6.4 PGA3 The finest gain and offset tuning is performed with the third and last PGA stage, according to the coding of Table 17-17 and Table 17-18. The output of PGA3 is also the input of the ADC. Thus, similarly to PGA2, we find that the voltage entering the ADC is given by: © Semtech 2006 www.semtech.com 17-10 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver V IN , ADC = GD3 ⋅ V D 2 − GDoff 3 ⋅ V REF (V) (Eq. 7) where GD3 and GDoff3 are respectively the gain and offset of PGA3 (in V/V). The control words are PGA3_GAIN[6:0] and PGA3_OFFSET[6:0] . To remain within the signal compliance of the PGA stages, the condition: V D1 , V D 2 < V DD (V) (Eq. 8) must be verified. Finally, combining equations Eq. 5 to Eq. 7 for the three PGA stages, the input voltage VIN,ADC of the ADC is related to VIN by: V IN , ADC = GDTOT ⋅ V IN − GDoff TOT ⋅ V REF (V) (Eq. 9) where the total PGA gain is defined as: GDTOT = GD3 ⋅ GD2 ⋅ GD1 (V/V) (Eq. 10) and the total PGA offset is: GDoff TOT = GDoff 3 + GD3 ⋅ GDoff 2 (V/V) (Eq. 11) 17.7 ADC Characteristics The main performance characteristics of the ADC (resolution, conversion time, etc.) are determined by three programmable parameters: sampling frequency fS, over-sampling ratio OSR, and number of elementary conversions NELCONV. The setting of these parameters and the resulting performances are described hereafter. 17.7.1 Conversion Sequence A conversion is started each time the bit START or the bit DEF is set. As depicted in Figure 17-5, a complete analog-to-digital conversion sequence is made of a set of NELCONV elementary incremental conversions and a final quantitative step. Each elementary conversion is made of (OSR+1) sampling periods TS=1/fS, i.e.: TELCONV = (OSR + 1) / f S (s) (Eq. 12) The result is the mean of the elementary conversion results. An important feature is that the elementary conversions are alternatively performed with the offset of the internal amplifiers contributing in one direction and the other to the output code. Thus, converter internal offset is eliminated if at least two elementary sequences are performed (i.e. if NELCONV ≥ 2). A few additional clock cycles are also required to initiate and end the conversion properly. © Semtech 2006 www.semtech.com 17-11 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver TELCONV= (OSR+1)/f S Init Elementary Conversion Conversion index Offset Elementary Conversion 1 + 2 - TCONV Elementary Conversion Elementary Conversion NELCONV- 1 + N ELCONV - End Conversion Result Figure 17-5. Analog-to-digital conversion sequence 17.7.2 Sampling Frequency The word FIN[1:0] is used to select the sampling frequency fS (Table 17-19). Three sub-multiples of the internal RC-based frequency fRCEXT can be chosen. For FIN = "11", sampling frequency is about 8kHz. Additional information on oscillators and their control can be found in the clock block documentation. FIN[1:0] 00 01 10 11 Sampling Frequency fS (Hz) XE8802 XE8801/05 1/4⋅fRC 1/8⋅fRCEXT 1/16⋅fRCEXT 1/8⋅fRC 1/64⋅fRCEXT 1/32⋅fRC ∼8kHz ∼4kHz Table 17-19. Sampling frequency settings (fRC= RC-based frequency) 17.7.3 Over-Sampling Ratio The over-sampling ratio (OSR) defines the number of integration cycles per elementary conversion. Its value is set with the word SET_OSR[2:0] in power of 2 steps (see Table 17-20) given by: 3 + SET_OSR[2 : 0] OSR = 2 (-) SET_OSR[2:0] (RegAcCfg0[4:2]) 000 001 010 011 100 101 110 111 (Eq. 13) Over-Sampling Ratio OSR (-) 8 16 32 64 128 256 512 1024 Table 17-20. Over-sampling ratio settings © Semtech 2006 www.semtech.com 17-12 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 17.7.4 Elementary Conversions As mentioned previously, the whole conversion sequence is made of a set of NELCONV elementary incremental conversions. This number is set with the word SET_NELC[1:0] in binary steps (see Table 17-21) given by: SET_NELC[1: 0] N ELCONV = 2 (-) (Eq. 14) # of Elementary Conversions NELCONV (-) 1 2 4 8 SET_NELC[1:0] (RegAcCfg0[6:5]) 00 01 10 11 Table 17-21. Number of elementary conversion settings As already mentioned, NELCONV must be equal or greater than 2 to reduce internal amplifier offsets. 17.7.5 Resolution The theoretical resolution of the ADC, without considering thermal noise, is given by: n = 2 ⋅ log 2 (OSR) + log 2 ( N ELCONV ) (Bits) (Eq. 15) 17 Resolution - n [Bits] 15 13 11 SET_NELC= 11 10 01 00 9 7 5 000 001 010 011 100 101 110 111 SET_OSR Figure 17-6. Resolution vs. SET_OSR[2:0] and SET_NELC[2:0] SET_OSR [2:0] 00 SET_NELC 01 10 11 000 001 010 011 100 101 110 111 6 8 10 12 14 16 16 16 7 9 11 13 15 16 16 16 9 11 13 15 16 16 16 16 8 10 12 14 16 16 16 16 (shaded area: resolution truncated to 16 bits due to output register size RegAcOut[15:0]) Table 17-22. Resolution vs. SET_OSR[2:0] and SET_NELC[1:0] settings © Semtech 2006 www.semtech.com 17-13 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Using Table 17-22 or the graph plotted in Figure 17-6, resolution can be set between 6 and 16 bits. Notice that, because of 16-bit register use for the ADC output, practical resolution is limited to 16 bits, i.e. n ≤ 16. Even if the resolution is truncated to 16 bit by the output register size, it may make sense to set OSR and NELCONV to higher values in order to reduce the influence of the thermal noise in the PGA (see section 17.8.4). 17.7.6 Conversion Time & Throughput As explained using Figure 17-5, conversion time is given by: TCONV = ( N ELCONV ⋅ (OSR + 1) + 1) / f S (s) (Eq. 16) and throughput is then simply 1/TCONV. For example, consider an over-sampling ratio of 256, 2 elementary conversions, and a sampling frequency of 500kHz (SET_OSR = "101", SET_NELC = "01", fRC = 2MHz, and FIN = "00"). In this case, using Table 17-23, the conversion time is 515 sampling periods, or 1.03ms. This corresponds to a throughput of 971Hz in continuous-time mode. The plot of Figure 17-7 illustrates the classic trade-off between resolution and conversion time. SET_OSR [2:0] 000 001 010 011 100 101 110 111 SET_NELC[1:0] 01 10 19 37 35 69 67 133 131 261 259 517 515 1029 1027 2053 2051 4101 00 10 18 34 66 130 258 514 1026 11 73 137 265 521 1033 2057 4105 8201 Table 17-23. Normalized conversion time (TCONV ⋅fS) vs. SET_OSR[2:0] and SET_NELC[1:0](normalized to sampling period 1/fS) Resolution - n [Bits] 16.0 14.0 12.0 10.0 8.0 6.0 10 00 4.0 10.0 11 SET_NELC 01 100.0 1000.0 10000.0 Normalized Conversion Time - TCONV*fS [-] Figure 17-7. Resolution vs. normalized conversion time for different SET_NELC[1:0] 17.7.7 Output Code Format The ADC output code is a 16-bit word in two's complement format (see Table 17-24). For input voltages outside the range, the output code is saturated to the closest full-scale value (i.e. 0x7FFF or 0x8000). For resolutions smaller than 16 bits, the non-significant bits are forced to the values shown in Table 17-25. The output code, expressed in LSBs, corresponds to: OUTADC = 216 ⋅ VIN , ADC OSR + 1 (LSB) ⋅ VREF OSR (Eq.17) © Semtech 2006 www.semtech.com 17-14 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Recalling equation Eq. 9, this can be rewritten as: OUTADC = 216 ⋅ VIN VREF ⎛ V ⋅ ⎜⎜ GDTOT − GDoff TOT ⋅ REF VIN ⎝ ⎞ OSR + 1 ⎟⎟ ⋅ OSR ⎠ (LSB) (Eq. 18) where, from Eq. 10 and Eq. 11, the total PGA gain and offset are respectively: GDTOT = GD3 ⋅ GD2 ⋅ GD1 (V/V) and: GDoff TOT = GDoff 3 + GD3 ⋅ GDoff 2 (V/V) ADC Input Voltage VIN,ADC % of Full Scale (FS) +2.49505V +0.5⋅FS +2.49497V ... ... +76.145µV 0V -76.145µV ... ... ... 0 ... ... -2.49505V ... -2.49513V -0.5⋅FS Output in LSBs Output Code in Hex 15 +2 -1 =+32'767 +215-2 =+32'766 ... +1 0 -1 ... 15 -2 -1 =-32'767 -215 =-32'768 7FFF 7FFE ... 0001 0000 FFFF ... 8001 8000 Table 17-24. Basic ADC Relationships (example for: VREF = 5V, OSR = 512, n = 16 bits) SET_OSR [2:0] 000 001 010 011 100 101 110 111 SET_NELC = 00 SET_NELC = 01 SET_NELC = 10 SET_NELC = 11 1000000000 10000000 100000 1000 10 - 100000000 1000000 10000 100 1 - 10000000 100000 1000 10 - 1000000 10000 100 1 - Table 17-25. Last forced LSBs in conversion output registers for resolution settings smaller than 16 bits (n < 16) (RegAcOutMsb[7:0] & RegAcOutLsb[7:0]) The equivalent LSB size at the input of the PGA chain is: LSB = 1 V REF OSR ⋅ ⋅ n 2 GDTOT OSR + 1 (V) (Eq. 19) Notice that the input voltage VIN,ADC of the ADC must satisfy the condition: © Semtech 2006 www.semtech.com 17-15 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver VIN , ADC ≤ 1 OSR ⋅ (V REFP − VREFN ) ⋅ 2 OSR + 1 (V) (Eq. 20) to remain within the ADC input range. 17.7.8 Power Saving Modes During low-speed operation, the bias current in the PGAs and ADC can be programmed to save power using the control words IB_AMP_PGA[1:0] and IB_AMP_ADC[1:0] (see Table 17-26). If the system is idle, the PGAs and ADC can even be disabled, thus, reducing power consumption to its minimum. This can considerably improve battery lifetime. ADC PGA [1:0] Bias Current Bias Current 00 01 10 11 x 1/4⋅IADC 1/2⋅IADC 3/4⋅IADC IADC x x 00 01 10 11 x 1/4⋅IPGA 1/2⋅IPGA 3/4⋅IPGA IPGA IB_AMP_ADC [1:0] IB_AMP_PGA Max. fS [kHz] 62.5 125 250 500 62.5 125 250 500 Table 17-26. ADC & PGA power saving modes and maximum sampling frequency 17.8 Specifications and Measured Curves This section presents measurement results for the acquisition chain. A summary table with circuit specifications and measured curves are given. 17.8.1 Default Settings Unless otherwise specified, the measurement conditions are the following: Temperature TA = +25°C VDD = +5V, GND = 0V, VREF = +5V, VIN = 0V RC frequency fRC = 2MHz, sampling frequency fS = 500kHz Offsets GDOff2 = GDOff3 = 0 Power operation: normal (IB_AMP_ADC[1:0] = IB_AMP_PGA[1:0] = '11') Resolution: for n = 12 bits: OSR = 32 and NELCONV = 4 o for n = 16 bits: OSR = 512 and NELCONV = 2 © Semtech 2006 www.semtech.com 17-16 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 17.8.2 Specifications Unless otherwise specified: Temperature TA = +25°C, VDD = +5V, GND = 0V, VREF = +5V, VIN = 0V, RC frequency fRC = 2MHz, sampling frequency fS = 500kHz, Overall PGA gain GDTOT = 1, offsets GDOff2 = GDOff3 = 0. Power operation: normal (IB_AMP_ADC[1:0] = IB_AMP_PGA[1:0] = '11'). For resolution n = 12 bits: OSR = 32 and NELCONV = 4. For resolution n = 16 bits: OSR = 512 and NELCONV = 2. VALUE PARAMETER UNITS COMMENTS/CONDITIONS MIN TYP MAX ANALOG INPUT CHARACTERISTICS +2.42 Gain = 1, OSR = 32 (Note 1) -2.42 V Differential Input Voltage Ranges +24.2 Gain = 100, OSR = 32 -24.2 mV VIN = (VINP - VINN) +2.42 Gain = 1000, OSR = 32 -2.42 mV Reference Voltage Range VREF = (VREFP – VREFN) PROGRAMMABLE GAIN AMPLIFIERS (PGA) Total PGA Gain, GDTOT PGA1 Gain, GD1 PGA2 Gain, GD2 PGA3 Gain, GD3 Gain Setting Precision (each stage) Gain Temperature Dependence Offset PGA2 Offset, GDoff2 PGA3 Offset, GDoff3 Offset Setting Precision (PGA2 or 3) Offset Temperature Dependence Input Impedance PGA1 PGA2, PGA3 Output RMS Noise PGA1 PGA2 PGA3 ADC STATIC PERFORMANCE Resolution, n No Missing Codes Gain Error Offset Error 0.5 1 1 0 -3 -1 -127/12 -3 Throughput Rate (Continuous Mode), 1/TCONV Nbr of Initialization Cycles, NINIT Nbr of End Conversion Cycles, NEND PGA Stabilization Delay DIGITAL OUTPUT ADC Output Data Coding ±0.5 ±5 V 1000 10 10 127/12 +3 V/V V/V V/V V/V % ppm/°C +1 +127/12 +3 V/V V/V % ppm/°C 1500 150 150 205 340 365 6 Integral Non-Linearity, INL Resolution n = 16 Bits Differential Non-Linearity, DNL Resolution n = 16 Bits Power Supply Rejection Ratio, PSRR DYNAMIC PERFORMANCE Sampling Frequency, fS Conversion Time, TCONV ±0.5 ±5 VDD 16 See Table 17-14 See Table 17-15 Step=1/12 V/V, See Table 17-17 Step=0.2 V/V, See Table 17-16 Step=1/12 V/V, See Table 17-18 (Note 2) kΩ kΩ kΩ PGA1 Gain = 1 (Note 3) PGA1 Gain = 10 (Note 3) Maximal gain (Note 3) µV µV µV (Note 4) (Note 5) (Note 6) Bits (Note 7) (Note 8) (Note 9) n = 16 bits (Note 10) ±0.15 ±1 % of FS LSB ±1.0 LSB (Note 11) ±0.5 78 72 LSB dB dB (Note 12) VDD = 5V ± 0.3V (Note 13) VDD = 3V ± 0.3V (Note 13) 3 133 1027 3.76 0.49 0 0 2 5 OSR kHz cycles/fS cycles/fS kSps kSps cycles cycles cycles n = 12 bits (Note 14) n = 16 bits (Note 14) n = 12 bits, fS = 500kHz n = 16 bits, fS = 500kHz (Note 15) Binary Two’s Complement See Table 17-24 and Table 17-25 © Semtech 2006 www.semtech.com 17-17 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Specifications (Cont’d) PARAMETER POWER SUPPLY Voltage Supply Range, VDD Analog Quiescent Current Consumption, Total (IQ) ADC Only PGA1 PGA2 PGA3 Analog Power Dissipation Normal Power Mode 3/4 Power Reduction Mode 1/2 Power Reduction Mode 1/4 Power Reduction Mode TEMPERATURE Specified Range Operating Range MIN VALUE TYP MAX +2.4 +5 +5.5 -40 -40 UNITS V 720/620 250/190 165/150 130/120 175/160 µA µA µA µA µA 3.6/1.9 2.7/1.4 1.8/0.9 0.9/0.5 mW mW mW mW +85 +125 COMMENTS/CONDITIONS Only Acquisition Chain VDD = 5V/3V VDD = 5V/3V VDD = 5V/3V VDD = 5V/3V VDD = 5V/3V All PGAs & ADC Active VDD = 5V/3V (Note 16) VDD = 5V/3V (Note 17) VDD = 5V/3V (Note 18) VDD = 5V/3V (Note 19) °C °C Notes: (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) = GD1⋅GD2⋅GD3. Maximum input voltage is given by: Gain defined as overall PGA gain GDTOT VIN,MAX = ±(VREF/2)⋅(OSR/OSR+1). Offset due to tolerance on GDoff2 or GDoff3 setting. For small intrinsic offset, use only ADC and PGA1. Measured with block connected to inputs through AMUX block. Normalized input sampling frequency for input impedance is fS = 512kHz. This figure must be multiplied by 2 for fS = 256kHz, 4 for fS = 128kHz. Input impedance is proportional to 1/fS. See model of Figure 17-18(a). Figure independent from PGA1 gain and sampling frequency fS. See equation Eq. 21 to calculate equivalent input noise. Figure independent on PGA2 gain and sampling frequency fS. See model of Figure 17-18(a). See equation Eq. 21 to calculate equivalent input noise. Figure independent on PGA3 gain and sampling frequency fS. See model of Figure 17-18(a) and equation Eq. 21 to calculate equivalent input noise. Resolution is given by n = 2⋅log2(OSR) + log2(NELCONV). OSR can be set between 8 and 1024, in powers of 2. NELCONV can be set to 1, 2, 4 or 8. If a ramp signal is applied to the input, all digital codes appear in the resulting ADC output data. Gain error is defined as the amount of deviation between the ideal (theoretical) transfer function and the measured transfer function (with the offset error removed). (See Figure 17-19) Offset error is defined as the output code error for a zero volt input (ideally, output code = 0). For ± 1 LSB offset, NELCONV must be ≥2. INL defined as the deviation of the DC transfer curve of each individual code from the best-fit straight line. This specification holds over the full scale. DNL is defined as the difference (in LSB) between the ideal (1 LSB) and measured code transitions for successive codes. Figures for Gains = 1 to 100. PSRR is defined as the amount of change in the ADC output value as the power supply voltage changes. Conversion time is given by: TCONV = (NELCONV ⋅ (OSR + 1) + 1) / fS. OSR can be set between 8 and 1024, in powers of 2. NELCONV can be set to 1, 2, 4 or 8. PGAs are reset after each writing operation to registers RegAcCfg1-5. The ADC must be started after a PGA or inputs commonmode stabilisation delay. This is done by writing bit Start several cycles after PGA settings modification or channel switching. Delay between PGA start or input channel switching and ADC start should be equivalent to OSR (between 8 and 1024) number of cycles. This delay does not apply to conversions made without the PGAs. Nominal (maximum) bias currents in PGAs and ADC, i.e. IB_AMP_PGA[1:0] = ‘11’ and IB_AMP_ADC[1:0] = ‘11’. Bias currents in PGAs and ADC set to 3/4 of nominal values, i.e. IB_AMP_PGA[1:0] = ‘10’, IB_AMP_ADC[1:0] = ‘10’. Bias currents in PGAs and ADC set to 1/2 of nominal values, i.e. IB_AMP_PGA[1:0] = ‘01’, IB_AMP_ADC[1:0] = ‘01’. Bias currents in PGAs and ADC set to 1/4 of nominal values, i.e. IB_AMP_PGA[1:0] = ‘00’, IB_AMP_ADC[1:0] = ‘00’. 17.8.3 Linearity 17.8.3.1 Integral non-linearity The integral non-linearity depends on the selected gain configuration. First of all, the non-linearity of the ADC (all PGA stages bypassed) is shown in Figure 17-8. © Semtech 2006 www.semtech.com 17-18 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Figure 17-8. Integral non-linearity of the ADC (PGA disabled, reference voltage of 4.8V) The different PGA stages have been designed to find the best compromise between the noise performance, the integral non-linearity and the power consumption. To obtain this, the first stage has the best noise performance and the third stage the best linearity performance. For large input signals (small PGA gains, i.e. up to about 50), the noise added by the PGA is very small with respect to the input signal and the second and third stage of the PGA should be used to get the best linearity. For small input signals (large gains, i.e. above 50), the noise level in the PGA is important and the first stage of the PGA should be used. The following figures give the non-linearity for different gain settings of the PGA, selecting the appropriate stage to get the best noise and linearity performance. Figure 17-9 shows the non-linearity when the third stage is used with a gain of 1. It is of course not very useful to use the PGA with a gain of 1 unless it is used to compensate offset. By increasing the gain, the integral non-linearity becomes even smaller since the signal in the amplifiers reduces. Figure 17-10 shows the non-linearity for a gain of 2. Figure 17-11 shows the non-linearity for a gain of 5. Figure 17-12 shows the non-linearity for a gain of 10. By comparing these figures to Figure 17-8, it can be seen that the third stage of the PGA does not add significant integral non-linearity. Figure 17-13 shows the non-linearity for a gain of 20 and Figure 17-14 shows the non-linearity for a gain of 50. In both cases the PGA2 is used at a gain of 10 and the remaining gain is realized by the third stage. It can be seen again that the second stage of the PGA does not add significant non-linearity. For gains above 50, the first stage PGA1 should be selected instead of PGA2. Although the non-linearity in the first stage of the PGA is larger than in stage 2 and 3, the gain in stage 3 is now sufficiently high so that the non-linearity of the first stage does become negligible as is shown in Figure 17-15 for a gain of 100. Therefore, the first stage is preferred over the second stage since it has less noise. Increasing the gain further up to 1000 will further increase the linearity since the signal becomes very small in the first two stages. The signal is full scale at the output of stage 3 and as shown in Figure 17-9 to Figure 17-12, this stage has very good linearity. © Semtech 2006 www.semtech.com 17-19 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Figure 17-9. Integral non-linearity of the ADC and with gain of 1 (PGA1 and PGA2 disabled, PGA3=1, reference voltage of 5V) Figure 17-10. Integral non-linearity of the ADC and gain of 2 (PGA1 and PGA2 disabled, PGA3=2 reference voltage of 5V) Figure 17-11. Integral non-linearity of the ADC and gain of 5 (PGA1 and PGA2 disabled, PGA3=5, reference voltage of 5V) © Semtech 2006 www.semtech.com 17-20 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Figure 17-12. Integral non-linearity of the ADC and gain of 10 (PGA1 and PGA2 disabled, PGA3=10, reference voltage of 5V) Figure 17-13. Integral non-linearity of the ADC and gain of 20 (PGA1 and PGA2=10, PGA3=2, reference voltage of 5V) © Semtech 2006 www.semtech.com 17-21 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Figure 17-14. Integral non-linearity of the ADC and gain of 50 (PGA1 disabled, PGA2=10, PGA3=5, reference voltage of 5V) Figure 17-15. Integral non-linearity of the ADC and gain of 100 (PGA1=10 and PGA3=10, PGA2 disabled, reference voltage of 5V) 17.8.3.2 Differential non-linearity The differential non-linearity is generated by the ADC. The PGA does not add differential non-linearity. Figure 17-16 shows the differential non-linearity. © Semtech 2006 www.semtech.com 17-22 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Figure 17-16. Differential non-linearity of the ADC converter. 17.8.4 Noise Ideally, a constant input voltage VIN should result in a constant output code. However, because of circuit noise, the output code may vary for a fixed input voltage. Thus, a statistical analysis on the output code of 1200 conversions for a constant input voltage was performed to derive the equivalent noise levels of PGA1, PGA2, and PGA3. The extracted rms output noise of PGA1, 2, and 3 are given in Table 17-27: standard output deviation and output rms noise voltage. Figure 17-17 shows the distribution for the ADC alone (PGA1, 2, and 3 bypassed). Quantitative noise is dominant in this case, and, thus, the ADC thermal noise is below 16 bits. The simple noise model of Figure 17-18(a) is used to estimate the equivalent input referred rms noise VN,IN of the acquisition chain in the model of Figure 17-18(b). This is given by the relationship: V N , IN = 2 (V N 1 / GD1 ) 2 + (V N 2 /(GD1 ⋅ GD2 )) 2 + (V N 3 /(GD1 ⋅ GD2 ⋅ GD3 )) 2 (OSR ⋅ N ELCONV ) (V2rms) (Eq. 21) where VN1, VN2, and VN3 are the output rms noise figures of Table 17-27, GD1, GD2, and GD3 are the PGA gains of stages 1 to 3 respectively. As shown in this equation, noise can be reduced by increasing OSR and NELCONV (increases the ADC averaging effect, but reduces noise). Parameter Standard deviation at ADC output (LSB) Output rms noise (µV) 1 PGA1 PGA2 PGA3 0.85 1.4 1.5 205 (VN1) 340 (VN2) 365 (VN3) Note: see noise model of Figure 17-18 and equation Eq. 21. Table 17-27. PGA noise measurements (n = 16 bits, OSR = 512, NELCONV = 2, VREF = 5V) © Semtech 2006 www.semtech.com 17-23 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Occurences [% of total samples] 80 60 40 20 0 -5 -4 -3 -2 -1 0 1 2 3 4 5 Output Code Deviation From Mean Value [LSB] Figure 17-17. ADC noise (PGA1, 2 & 3 bypassed, OSR=512,NELCONV=2) fS PGA1 VN1 GD1 PGA2 VN2 GD2 PGA3 VN3 GD3 ADC (a) fS PGA1 PGA2 PGA3 GD1 GD2 GD3 VN,IN ADC (b) Figure 17-18. (a) Simple noise model for PGAs and ADC and (b) total input referred noise As an example, consider the system where: GD2 = 10 (GD1 = 1; PGA3 bypassed), OSR = 512, NELCONV = 2, VREF = 5V. In this case, the noise contribution VN1 of PGA1 is dominant over that of PGA2. Using equation Eq. 21, we get: VN,IN = 6.4µV (rms) at the input of the acquisition chain, or, equivalently, 0.85 LSB at the output of the ADC. Considering a 0.2V (rms) maximum signal amplitude, the signal-to-noise ratio is 90dB. Implementing a software filter can also reduce noise. By taking an average on a number of subsequent measurements, the apparent noise is reduced by the square root of the number of measurement used to make the average. 17.8.5 Gain Error and Offset Error Gain error is defined as the amount of deviation between the ideal transfer function (theoretical equation Eq. 18) and the measured transfer function (with the offset error removed). The actual gain of the different stages can vary relating to the fabrication tolerances of the different elements. Although these tolerances are specified to a maximum of ±3%, most of the time they will be around ±0.5%. Moreover, the tolerances between the different stages are not correlated and the probability of getting the maximal error in the same direction in all stages is very low. Finally, the software can calibrate these gain errors at the same time as the gain errors of the sensor for instance. Figure 17-19 shows gain error drift vs. temperature for different PGA gains. The curves are expressed in % of FullScale Range (FSR) normalized to 25°C. © Semtech 2006 www.semtech.com 17-24 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Offset error is defined as the output code error for a zero volt input (ideally, output code = 0). The offset of the ADC and the PGA1 stage are completely suppressed if NELCONV > 1. The measured offset drift vs. temperature curves for different PGA gains are depicted in Figure 17-20. The output offset error, expressed in LSB for 16-bit setting, is normalized to 25°C. Notice that if the ADC is used alone, the output offset error is below ±1 LSB and has no drift. NORMALIZED TO 25°C Gain Error [% of FSR] 0.2 0.1 0.0 -0.1 1 5 20 100 -0.2 -0.3 -0.4 -50 -25 0 25 50 75 100 Temperature [°C] Figure 17-19. Gain error vs. temperature for different PGA gains Output Offset Error [LSB] NORMALIZED TO 25°C 100 1 5 20 100 80 60 40 20 0 -20 -40 -50 -25 0 25 50 75 100 Temperature [°C] Figure 17-20. Offset error vs. temperature for different PGA gains 17.8.6 Power Consumption Figure 17-21 plots the variation of quiescent current consumption with supply voltage VDD, as well as the distribution between the 3 PGA stages and the ADC (see Table 17-28). As shown in Figure 17-22, if lower sampling frequency is used, the quiescent current consumption can be lowered by reducing the bias currents of the PGAs and the ADC with registers IB_AMP_PGA [1:0] and IB_AMP_ADC [1:0]. (In Figure 17-22, IB_AMP_PGA/ADC[1:0] = '11', '10', '00' for fS = 500, 250, 62.5kHz respectively.) Quiescent current consumption vs. temperature is depicted in Figure 17-23, showing a relative increase of nearly 40% between -45 and +85°C. Figure 17-24 shows the variation of quiescent current consumption for different frequency settings of the internal RC oscillator. It can be seen that the quiescent current varies by about 20% between 100kHz and 2MHz. © Semtech 2006 www.semtech.com 17-25 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 800 Quiescent Current - IQ [µ A] 700 PGA1, 2 & 3 + ADC 600 500 PGA1 & 2 + ADC 400 PGA1 + ADC 300 200 No PGAs, ADC only 100 2.5 3.0 3.5 4.0 4.5 5.0 5.5 Supply Voltage - VDDA [V] Figure 17-21. Quiescent current consumption vs. supply voltage 800 Sampling Frequency fS : 500kHz Quiescent Current - IQ [µ A] 700 600 250kHz 500 400 300 62.5kHz 200 100 2.5 3.0 3.5 4.0 4.5 5.0 5.5 Supply Voltage - VDDA [V] Figure 17-22. Quiescent current consumption vs. supply voltage for different sampling frequencies 20 Relative Quiescent Current Change IQ / IQ,25°C [%] Quiescent Current - IQ [µ A] 900 850 800 750 700 650 600 550 15 10 5 0 -5 -10 -15 -20 -25 500 -50 -25 0 25 50 75 100 125 -50 -25 0 25 50 75 100 125 Temperature [°C] Temperature [°C] (a) (b) Figure 17-23. (a) Absolute and (b) relative change inquiescent current consumption vs. temperature © Semtech 2006 www.semtech.com 17-26 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Supply VDD = 5V VDD = 3V ADC PGA1 PGA2 PGA3 TOTAL Unit 250 190 165 150 130 120 175 160 720 620 µA µA Table 17-28. Typical quiescent current distributions in acquisition chain 15 850 10 800 Quiescent Current - IQ [µ A] Relative Quiescent Current Change ∆ IQ / IQ,2MHz [%] (n = 16 bits, fS = 500kHz) 5 0 -5 -10 -15 750 700 650 600 550 500 -20 0 500 1000 1500 2000 2500 3000 0 3500 500 1000 1500 2000 2500 3000 3500 Frequency - fRC [kHz] Frequency - fRC [kHz] (a) (b) Figure 17-24. (a) Absolute and (b) relative change in quiescent curent consumption vs. RC oscillator frequency (all PGAs active, VDD = 5V) 17.8.7 Power Supply Rejection Ratio Figure 17-25 shows power supply rejection ratio (PSRR) at a 3V and a 5V supply voltage, and for various PGA gains. PSRR is defined as the ratio (in dB) of voltage supply change (in V) to the change in the converter output (in V). PSRR depends on both PGA gain and supply voltage VDD. 105 VDD=3V VDD=5V 100 PSRR [dB] 95 90 85 80 75 70 65 60 1 5 10 20 100 PGA Gain [V/V] Figure 17-25. Power supply rejection ratio (PSRR) Supply VDD = 5V VDD = 3V GAIN = 1 GAIN =5 GAIN = 10 GAIN = 20 GAIN =100 Unit 79 72 78 79 100 90 99 90 97 86 dB dB Table 17-29. PSRR (n = 16 bits, VIN = VREF = 2.5V, fS = 500kHz) © Semtech 2006 www.semtech.com 17-27 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 17.9 Application Hints 17.9.1 Input Impedance The PGAs of the acquisition chain employ switched-capacitor techniques. For this reason, while a conversion is done, the input impedance on the selected channel of the PGAs is inversely proportional to the sampling frequency fS and to stage gain as given in equation 22. Z in ≥ 768 ⋅ 10 9 ΩHz f s ⋅ gain (Eq. 22) The input impedance observed is the input impedance of the first PGA stage that is enabled or the input impedance of the ADC if all three stages are disabled. PGA1 (with a gain of 10), PGA2 (with a gain of 10) and PGA3 (with a gain of 10) each have a minimum input impedance of 150kΩ at fS = 512kHz (see Specification Table). Larger input impedance can be obtained by reducing the gain and/or by reducing the sampling frequency. Therefore, with a gain of 1 and a sampling frequency of 100kHz, Zin > 7.6MΩ. The input impedance on channels that are not selected is very high (>100MΩ). 17.9.2 PGA Settling or Input Channel Modifications PGAs are reset after each writing operation to registers RegAcCfg1-5. Similarly, input channels are switched after modifications of AMUX[4:0] or VMUX. To ensure precise conversion, the ADC must be started after a PGA or inputs common-mode stabilization delay. This is done by writing bit START several cycles after modification of PGA settings or channel switching. Delay between PGA start or input channel switching and ADC start should be equivalent to OSR (between 8 and 1024) number of cycles. This delay does not apply to conversions made without the PGAs. If the ADC is not settled within the specified period, there is probably an input impedance problem (see previous section). 17.9.3 PGA Gain & Offset, Linearity and Noise Hereafter are a few design guidelines that should be taken into account when using the ZoomingADC™: 1) 2) 3) 4) 4) 5) 6) Keep in mind that increasing the overall PGA gain, or "zooming" coefficient, improves linearity but degrades noise performance. Use the minimum number of PGA stages necessary to produce the desired gain ("zooming") and offset. Bypass unnecessary PGAs. For high gains (>50), use PGA stage 1. For low gains (<50) use stages 2 and 3. For the lowest noise, set the highest possible gain on the first (front) PGA stage used in the chain. For example, in an application where a gain of 20 is needed, set the gain of PGA2 to 10, set the gain of PGA3 to 2. For highest linearity and lowest noise performance, bypass all PGAs and use the ADC alone (applications where no "zooming" is needed); i.e. set ENABLE[3:0] = '0001'. For low-noise applications where power consumption is not a primary concern, maintain the largest bias currents in the PGAs and in the ADC; i.e. set IB_AMP_PGA[1:0] = IB_AMP_ADC[1:0] = '11'. For lowest output offset error at the output of the ADC, bypass PGA2 and PGA3. Indeed, PGA2 and PGA3 typically introduce an offset of about 5 to 10 LSB (16 bit) at their output. Note, however, that the ADC output offset is easily calibrated out by software. © Semtech 2006 www.semtech.com 17-28 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 17.9.4 Frequency Response The incremental ADC is an over-sampled converter with two main blocks: an analog modulator and a low-pass digital filter. The main function of the digital filter is to remove the quantitative noise introduced by the modulator. As shown in Figure 17-26, this filter determines the frequency response of the transfer function between the output of the ADC and the analog input VIN. Notice that the frequency axes are normalized to one elementary conversion period OSR/fS. The plots of Figure 17-26 also show that the frequency response changes with the number of elementary conversions NELCONV performed. In particular, notches appear for NELCONV ≥ 2. These notches occur at: f NOTCH (i) = i ⋅ fS OSR ⋅ N ELCONV for i = 1,2,..., ( N ELCONV − 1) (Hz) (Eq. 23) and are repeated every fS/OSR. 1.2 1 Normalized Magnitude [-] Normalized Magnitude [-] Information on the location of these notches is particularly useful when specific frequencies must be filtered out by the acquisition system. For example, consider a 5Hz-bandwidth, 16-bit sensing system where 50Hz line rejection is needed. Using the above equation and the plots below, we set the 4th notch for NELCONV = 4 to 50Hz, i.e. 1.25⋅fS/OSR = 50Hz. The sampling frequency is then calculated as fS = 20.48kHz for OSR = 512. Notice that this choice yields also good attenuation of 50Hz harmonics. NELCONV = 1 0.8 0.6 0.4 0.2 0 1.2 1 N ELCONV = 2 0.8 0.6 0.4 0.2 0 0 1 2 3 4 0 1.2 1 NELCONV = 4 0.8 0.6 0.4 0.2 0 0 1 2 3 1 2 3 4 Normalized Frequency - f *(OSR/fS) [-] Normalized Magnitude [-] Normalized Magnitude [-] Normalized Frequency - f *(OSR/fS) [-] 4 1.2 N ELCONV = 8 1 0.8 0.6 0.4 0.2 0 0 1 2 3 4 Normalized Frequency - f *(OSR/fS) [-] Normalized Frequency - f *(OSR/fS) [-] Figure 17-26. Frequency response: normalized magnitude vs. frequency for different NELCONV 17.9.5 Power Reduction The ZoominADC™ is particularly well suited for low-power applications. When very low power consumption is of primary concern, such as in battery operated systems, several parameters can be used to reduce power consumption as follows: 1) 2) 3) Operate the acquisition chain with a reduced supply voltage VDD. Disable the PGAs which are not used during analog-to-digital conversion with ENABLE[3:0]. Disable all PGAs and the ADC when the system is idle and no conversion is performed. © Semtech 2006 www.semtech.com 17-29 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 4) 5) Use lower bias currents in the PGAs and the ADC using the control words IB_AMP_PGA[1:0] and IB_AMP_ADC[1:0]. (This reduces the maximum sampling frequency according to Table 17-26.) Reduce internal RC oscillator frequency and/or sampling frequency. Finally, remember that power reduction is typically traded off with reduced linearity, larger noise and slower maximum sampling speed. © Semtech 2006 www.semtech.com 17-30 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 18. 18.1 18.2 18.3 Voltage Multiplier FEATURES OVERVIEW CONTROL PART REGISTERS 18-2 18-2 18-2 © Semtech 2006 www.semtech.com 18-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 18.1 Features • • Generates a voltage that is higher or equal to the supply voltage, but at maximum 4.8 V Can be easily enabled or disabled 18.2 Overview en_vmult switch power regs switched analog (other blocks) VMULT ck_vmult (analog) prescaler Vmult (capacitor only required when enabled) Vmult_dl Figure 17-1 : Structure of the Vmult peripheral The Vmult block generates a voltage (called “Vmult”) that is higher or equal to the supply voltage. This output voltage is intended for use in analog switch drivers, for example in the ADC and PGA block. • In normal use its value is 4.8 V, this voltage is available on the pad Vmult. • When the main voltage is below 2.6 V, Vmult is twice the main voltage minus 0.4 V. • When the main voltage is above 4.8 V, Vmult remains 4.8 V but the internal voltage intended for the analog switch drivers equal the main voltage. The voltage multiplier should be on (bit Enable in RegVmultCfg0) when using switched analog blocks, like ADC, DAC or analog properties of Port B when Vbat is below 3V. The source clock of Vmult is selected from Fin[1:0] in RegVmultCfg0. It is strongly recommended to use the same Fin bit code as in the ADC. The external capacitor on the pin VMULT has to be connected if the block is enabled. The size of the capacitor has to be 2nF ± 50%. 18.3 Control part Registers There is only one register in the Vmult. Table 18-1 describes the bits in the register. Pos. 2 RegVmultCfg0 Enable rw rw Reset 0 1-0 Fin rw 0 Function enable of the vmult ‘1’ : enabled ‘0’ : disabled system clock division factor ‘00’ : 1/2, ‘01’ : 1/4, ‘10’ : 1/16, ‘11’ : 1/64 Table 18-1. RegVmultCfg0 © Semtech 2006 www.semtech.com 18-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 19. 19.1 19.2 19.3 19.4 19.4.1 19.4.2 19.4.3 19.4.4 19.5 19.5.1 19.5.2 19.5.3 19.5.4 19.6 19.6.1 19.6.2 19.7 19.7.1 19.7.2 19.8 19.8.1 19.8.2 19.8.3 19.8.4 LCD Driver FEATURES ......................................................................................................................................... 19-2 OVERVIEW......................................................................................................................................... 19-2 REGISTER MAP .................................................................................................................................. 19-3 BASIC LCD CAPABILITIES .................................................................................................................... 19-6 Direct Drive Mode ............................................................................................................................ 19-7 1:2 Multiplex scheme ....................................................................................................................... 19-9 1:3 Multiplex scheme ..................................................................................................................... 19-11 1:4 Multiplex scheme ..................................................................................................................... 19-13 ADVANCED LCD FEATURES .............................................................................................................. 19-15 Register usage .............................................................................................................................. 19-15 Sleep mode or blinking mode ........................................................................................................ 19-15 LCD frame frequency selection...................................................................................................... 19-16 Voltage generation......................................................................................................................... 19-17 PARALLEL I/O PORT CAPABILITIES...................................................................................................... 19-21 Parallel port function ...................................................................................................................... 19-21 parallel port voltage ....................................................................................................................... 19-22 PARTIAL LCD DRIVER / PARTIAL PARALLEL I/O PORT............................................................................ 19-24 Single low impedance voltage for V3 of LCD and digital I/O .......................................................... 19-24 Different voltage for V3 of LCD and digital I/O ............................................................................... 19-26 SPECIFICATIONS .............................................................................................................................. 19-27 pad_lcd_io used in LCD mode....................................................................................................... 19-27 pad_lcd_io used in digital I/O mode ............................................................................................... 19-27 voltage reference ........................................................................................................................... 19-28 LCD multiplier/divider..................................................................................................................... 19-28 © Semtech 2006 www.semtech.com 19-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 19.1 Features • • • • • • • • • Supports direct drive for 32 segments. Supports multiplexed drive for up to 120 segments Multiplex 1/2, 1/3 and 1/4. Possibility to use pads as an input/output port or as LCD driver pin. Multiple frame frequencies. Sleep mode. On chip low-power voltage generation. LCD driver voltage is independent from the circuit supply voltage. Parallel IO with voltage different from the circuit supply voltage. 19.2 Overview S E G ME N TS W AV E F O R M G E N E R AT O R L C D P E R IP H D IS P L AY MEMORY 1 6 x 8 b it s COMS W AV E F O R M G E N E R AT O R LC D PAD S ts I/ O P O R T I/ O p o rt c o n t r o l r e g is t e rs LCD C O N TRO L AN D M O D E R E G IS T E R V1 VG E N CO NTR O L R E G IS T E R V2 V3 VG E N PA D S Figure 19-1: General Structure The LCD driver generates all waveforms to drive the display. The user has only to set a 0 (segment off) or a 1 (segment on) in the bit location corresponding to the segment. The LCD driver supports the wave form generation for different multiplexing schemes: direct drive (no multiplexing), multiplexing by 2, by 3 and by 4. The frame frequency is software programmable. Low power on-chip voltage generation is provided for each of the different multiplexing schemes. The LCD driver voltage is completely independent from the circuit supply voltage: the LCD driver voltage can be below or above the circuit supply voltage. All or part of the driver can be configured as a © Semtech 2006 www.semtech.com 19-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver general-purpose parallel IO-port. When used as a parallel port, it can be connected to circuits with a different supply voltage. Section 19.4 describes the basic functions of the LCD driver using the on-chip voltage generator. Section 19.5 describes more advanced the features of the LCD driver and especially different ways of generating the voltage for the LCD. Section 19.6 describes how to use the peripheral as a general-purpose parallel IO-port. Section 19.7 shows how to partition when used as a partial LCD driver and partial IO port. Finally, the electrical specifications of the driver are given in section 19.8. 19.3 Register map There are thirty-six registers in the LCD driver, namely RegVgenCfg0, RegLcdOn, RegLcdSe, RegLcdClkFrame, RegLcdDataN (0≤N≤15), RegPLcdInN (0≤N≤3), RegPLcdOutN (0≤N≤3), RegPLcdDirN (0≤N≤3) and RegPLcdPullupN (0≤N≤3). Table 19-2 to Table 19-11 show the mapping of control bits and functionality of these registers while Table 19-1 gives the default address of these thirty-six registers. Register name Vgen registers RegVgenCfg0 Control registers RegLcdOn RegLcdSe RegLcdClkFrame LCD registers RegLcdData0 RegLcdData1 RegLcdData2 RegLcdData3 RegLcdData4 RegLcdData5 RegLcdData6 RegLcdData7 RegLcdData8 RegLcdData9 RegLcdData10 RegLcdData11 RegLcdData12 RegLcdData13 RegLcdData14 RegLcdData15 Port Registers RegPLcdOut0 RegPLcdOut1 RegPLcdOut2 RegPLcdOut3 RegPLcdDir0 RegPLcdDir1 RegPLcdDir2 RegPLcdDir3 RegPLcdPull0 RegPLcdPull1 RegPLcdPull2 RegPLcdPull3 RegPLcdIn0 RegPLcdIn1 RegPLcdIn2 RegPLcdIn3 Table 19-1: LCD registers default addresses © Semtech 2006 www.semtech.com 19-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. RegVgenCfg0 RW reset 7-6 5-4 -VgenClkSel r rw 00 10 nresetglobal 3 VgenOff rw 1 nresetglobal 2 VgenMode rw 0 nresetglobal 1 0 VgenStdb VgenRefEn rw rw 0 nresetglobal 0 nresetglobal function Unused Voltage generator frequency selection. 00 – 256 Hz 01 – 512 Hz 10 – 1 kHz 11 – 2 kHz Voltage generator disable signal. 0 – enabled 1 – disabled Mode selection. 0 – 1/3 bias 1 – 1/2 bias Stand-by mode Internal 1.2V voltage reference enable. 0 – disabled 1 – enabled Table 19-2: RegVgenCfg0 pos. 7-3 2 RegLcdOn LcdSleep Rw rw Reset 00000 1 nresetglobal 1-0 Function LcdMux rw 00 nresetglobal 1 = stop operating 0 = continue operating 00 = direct drive 01 = 1:2 mux drive 10 = 1:3 mux drive 11 = 1:4 mux drive Table 19-3: RegLcdOn pos. 7 RegLcdSe LcdSe3 rw rw Reset 1 nresetglobal 6 LcdSe7 rw 1 nresetglobal 5 LcdSe11 rw 1 nresetglobal 4 LcdSe15 rw 1 nresetglobal 3 LcdSe19 rw 1 nresetglobal 2 LcdSe23 rw 1 nresetglobal 1 LcdSe27 rw 1 nresetglobal 0 LcdSe31 rw 1 nresetglobal Function 1 = pins pad_lcd_io(0-3) have LCD drive function 0 = pins pad_lcd_io(0-3) have digital function 1 = pins pad_lcd_io(4-7) have LCD drive function 0 = pins pad_lcd_io(4-7) have digital function 1 = pins pad_lcd_io(8-11) have LCD drive function 0 = pins pad_lcd_io(8-11) have digital function 1 = pins pad_lcd_io(12-15) have LCD drive function 0 = pins pad_lcd_io(12-15)have digital function 1 = pins pad_lcd_io(16-19) have LCD drive function 0 = pins pad_lcd_io(16-19) have digital function 1 = pins pad_lcd_io(20-23) have LCD drive function 0 = pins pad_lcd_io(20-23) have digital function 1 = pins pad_lcd_io(24-27) have LCD drive function 0 = pins pad_lcd_io(24-27) have digital function 1 = pins pad_lcd_io(28-31) have LCD drive function 0 = pins pad_lcd_io(28-31) have digital function Table 19-4: RegLcdSe © Semtech 2006 www.semtech.com 19-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7-5 RegLcdClkFrame LcdDivFreq rw rw reset 000 nresetglobal 4-2 1-0 Reserved LcdFreq rw 000 00 nresetglobal Function 000 = LcdFreq 001 = LcdFreq / 2 010 = LcdFreq / 3 011 = LcdFreq / 4 100 = LcdFreq / 5 101 = LcdFreq / 6 110 = LcdFreq / 7 111 = LcdFreq / 8 00 = 512 Hz 01 = 1024 Hz 10 = 2048 Hz 11 = 4096 Hz Table 19-5: RegLcdClkFrame Pos. 7 6 5 4 3 2 1 0 RegLcdDataN LcdDataN[7] LcdDataN[6] LcdDataN[5] LcdDataN[4] LcdDataN[3] LcdDataN[2] LcdDataN[1] LcdDataN[0] Rw rw rw rw rw rw rw rw rw Reset 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf Description Segment[3] value in 1:4 MUX Segment[2] value in 1:3 MUX Segment[1] value in 1:2 MUX Segment[0] value in DD Segment[3] value in 1:4 MUX Segment[2] value in 1:3 MUX Segment[1] value in 1:2 MUX Segment[0] value in DD Map Pin pad_lcd_io[2N+1] pad_lcd_io[2N] Table 19-6: RegLcdDataN with 0≤N≤14 Pos. 7 6 5 4 3 2 1 0 RegLcdData15 LcdData15[7] LcdData15[6] LcdData15[5] LcdData15[4] LcdData15[3] LcdData15[2] LcdData15[1] LcdData15[0] Rw rw rw rw rw rw rw rw rw Reset 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf Description Map Pin --Segment[1] value in 1:2 MUX Segment[0] value in DD -Segment[2] value in 1:3 MUX Segment[1] value in 1:2 MUX Segment[0] value in DD pad_lcd_io[31] pad_lcd_io[30] Table 19-7: RegLcdData15 Pos. 7 6 5 4 3 2 1 0 RegPLcdInN PLcdInN[7] PLcdInN[6] PLcdInN[5] PLcdInN[4] PLcdInN[3] PLcdInN[2] PLcdInN[1] PLcdInN[0] Rw r r r r r r r r Reset Description x x x x x x x x pad_lcd_io[8N+7] input value pad_lcd_io[8N+6] input value pad_lcd_io[8N+5] input value pad_lcd_io[8N+4] input value pad_lcd_io[8N+3] input value pad_lcd_io[8N+2] input value pad_lcd_io[8N+1] input value pad_lcd_io[8N+0] input value Table 19-8: RegPLcdInN with 0≤N≤3 © Semtech 2006 www.semtech.com 19-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Pos. 7 6 5 4 3 2 1 0 RegPLcdOutN PLcdOuN[7] PLcdOutN[6] PLcdOutN[5] PLcdOutN[4] PLcdOutN[3] PLcdOutN[2] PLcdOutN[1] PLcdOutN[0] Rw rw rw rw rw rw rw rw rw Reset 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf Description pad_lcd_io[8N+7] output value pad_lcd_io[8N+6] output value pad_lcd_io[8N+5] output value pad_lcd_io[8N+4] output value pad_lcd_io[8N+3] output value pad_lcd_io[8N+2] output value pad_lcd_io[8N+1] output value pad_lcd_io[8N+0] output value Table 19-9: RegPLcdOutN with 0≤N≤3 Pos. 7 6 5 4 3 2 1 0 RegPLcdDirN PLcdDirN[7] PLcdDirN[6] PLcdDirN[5] PLcdDirN[4] PLcdDirN[3] PLcdDirN[2] PLcdDirN[1] PLcdDirN[0] Rw rw rw rw rw rw rw rw rw Reset 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf Description pad_lcd_io[8N+7] direction (0=input) pad_lcd_io[8N+6] direction (0=input) pad_lcd_io[8N+5] direction (0=input) pad_lcd_io[8N+4] direction (0=input) pad_lcd_io[8N+3] direction (0=input) pad_lcd_io[8N+2] direction (0=input) pad_lcd_io[8N+1] direction (0=input) pad_lcd_io[8N+0] direction (0=input) Table 19-10: RegPLcdDirN with 0≤N≤3 Pos. 7 6 5 4 3 2 1 0 RegPLcdPullupN PLcdPullupN[7] PLcdPullupN[6] PLcdPullupN[5] PLcdPullupN[4] PLcdPullupN[3] PLcdPullupN[2] PLcdPullupN[1] PLcdPullupN[0] Rw rw rw rw rw rw rw rw rw Reset 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf 0 nresetpconf Description pullup for pad_lcd_io[8N+7] (1=active) pullup for pad_lcd_io[8N+6] (1=active) pullup for pad_lcd_io[8N+5] (1=active) pullup for pad_lcd_io[8N+4] (1=active) pullup for pad_lcd_io[8N+3] (1=active) pullup for pad_lcd_io[8N+2] (1=active) pullup for pad_lcd_io[8N+1] (1=active) pullup for pad_lcd_io[8N+0] (1=active) Table 19-11: RegPLcdPullupN with 0≤N≤3 19.4 Basic LCD capabilities This section shows how to use the LCD driver. For each multiplexing configuration it gives a basic example explaining how to set-up the driver, the generated waveforms, how to connect the display and how to use the onchip voltage generator. Other connection schemes and more advanced control functions will be given in the next section. The LCD module generates the timing control to drive a static or multiplexed LCD panel, with support for up to 32 segment lines multiplexed with up to four common lines. The table below summarises the multiplexing scheme for LCD operation. Multiplex scheme direct drive (DD) 1:2 1:3 1:4 Max #segments 32 32 x 2 31 x 3 30 x 4 Table 19-12: Multiplexing Scheme © Semtech 2006 www.semtech.com 19-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 19.4.1 Direct Drive Mode With direct drive mode, each pad_lcd_io pin drives one segment of the display. In this mode, up to 32 segments of the display can be connected directly to the pad_lcd_io[31:0] pins. The common (or backplane) node is to be connected to pad_lcd_com[0]. This connection is shown in Figure 19-2. pad_vgen_vb pad_lcd_vr1 pad_lcd_vr2 pad_vgen_v3 pad_vgen_v2 1.2V voltage reference pad_vgen_v1 C aux Clcd1 VGEN Clcd2 Clcd3 pad_vgen_va pad_lcd_io[31:0] SEGMENTS LCD DISPLAY pad_lcd_com1 pad_lcd_com0 COM0 Figure 19-2: LCD with direct drive and on-chip voltage generation. Lots of different configurations for the voltage generation can be used. Figure 19-2 shows the connections and external elements that are required if the display is to be driven from V3=3.6V independently from the circuit supply voltage. Other possible configurations are given in section 19.5.4. The recommended value for the external elements is 470nF (see also section 19.5.4.1). To operate the driver in this configuration, the registers should be loaded with the values in Table 19-13. Register RegVgenCfg0 RegLcdOn RegLcdSe RegLcdClkFrame Contents[7:0] xx110001 xxxxx000 11111111 100xxx00 Table 19-13. Register contents for direct drive example © Semtech 2006 www.semtech.com 19-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver V3 pad_lcd_com0 pad_lcd_com0 - pad_lcd_io[0] (off) V0 V3 V0 pad_lcd_io[1] pad_lcd_com0 - pad_lcd_io[1] (on) V0 V3 V3 V0 pad_lcd_io[0] -V3 V0 Figure 19-3: Direct Drive mode waveforms The voltage generator has to be configured for the use of the internal reference (VgenRefEn=1 in RegVgenCfg0). The generator has to run (VgenOff=0 and VgenStdb=0 in RegVgenCfg0). It has to multiply by 3 (VgenMode=0 in RegVgenCfg0) and the generator output impedance is set to minimum (VgenClkSel=11 in RegVgenCfg0). The LCD driver is set to direct drive (LcdMux=00 in RegLcdOn) and the waveform generator is enabled (LcdSleep=0 in RegLcdOn). All pads are set into the LCD driver mode by setting all bits of register RegLcdSe to 1. To select a frame rate of about 50Hz, set the bits LcdFreq=00 and LcdDivFreq=100 in RegLcdClkFrame. Note that the precision of the frame frequency depends on the selected clock source (see clock block documentation). The 32 segments are on or off depending on the bits set in the registers RegLcdDataN. Only the bits 0 and 4 of these registers are used in direct drive. All other bits are don’t care. Figure 19-3 shows the generated waveforms for two segments. Segment0 is off (LcdData0[0]=0 in RegLcdData0) and segment1 is on (LcdData0[4]=1 in RegLcdData0). The figure shows on the left side the waveforms on the circuit pins, in the middle the segment status and on the right the voltage on the segment (difference between the pad_lcd_io pin and pad_lcd_com0 pin). © Semtech 2006 www.semtech.com 19-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 19.4.2 1:2 Multiplex scheme With the 1:2 multiplex scheme, each pad_lcd_io pin drives two segments of the display. The segments of the display are connected to the pad_lcd_io[31:0] pins. The common (or backplane) nodes are to be connected to pad_lcd_com0 and pad_lcd_com1. This connection is shown in Figure 19-4. pad_vgen_vb pad_lcd_vr1 pad_lcd_vr2 pad_vgen_v3 pad_vgen_v2 1.2V voltage reference pad_vgen_v1 C aux Clcd1 VGEN Clcd2 pad_vgen_va pad_lcd_io[31:0] pad_lcd_com1 pad_lcd_com0 LCD DISPLAY SEGMENTS COM1 COM0 Figure 19-4. LCD with 1:2 multiplexing and on-chip 1/2 bias voltage generation. Lots of different configurations for the voltage generation can be used. Figure 19-4 shows the connections and external elements that are required if the display is to be driven from V2=1.2V and V3=2.4V independently from the circuit supply voltage. Other possible configurations are given in section 19.5.4. Recommended values for the external capacitors are 470nF. To operate the driver in this configuration, the registers should be loaded with the values in Table 19-14. Register RegVgenCfg0 RegLcdOn RegLcdSe RegLcdClkFrame Contents[7:0] xx110101 xxxxx001 11111111 100xxx01 Table 19-14. Register contents for 1:2 mux example The voltage generator has to be configured for the use of the internal reference (VgenRefEn=1 in RegVgenCfg0). The generator has to run (VgenOff=0 and VgenStdb=0 in RegVgenCfg0). It has to be in 1/2 bias (VgenMode=1 in RegVgenCfg0) and the generator output impedance is set to minimum (VgenClkSel=11 in RegVgenCfg0). © Semtech 2006 www.semtech.com 19-9 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver V3 pad_lcd_com0 V2 V0 V3 pad_lcd_com1 pad_lcd_com0 - pad_lcd_io[0] (on) V2 V3 V0 V2 V0 -V2 -V3 V3 pad_lcd_io[0] pad_lcd_com1 - pad_lcd_io[1] (off) V2 V3 V0 V2 V0 -V2 -V3 V3 pad_lcd_io[1] V2 V0 Figure 19-5. 1:2 MUX mode waveforms The LCD driver is set to 1:2 multiplexing (LcdMux=01 in RegLcdOn) and the waveform generator is enabled (LcdSleep=0 in RegLcdOn). All pads are set into the LCD driver mode by setting all bits of register RegLcdSe to 1. To select a frame rate of about 50Hz, set the bits LcdFreq=01 and LcdDivFreq=100 in RegLcdClkFrame. Note that the precision of the frame frequency depends on the selected clock source (see clock block documentation). The 64 segments are on or off depending on the bits set in the registers RegLcdDataN. Only the bits 0,1 and 4,5 of these registers are used in 1:2 multiplexing. Figure 19-5 shows the generated waveforms for four segments. Segments connected to pad_lcd_io[0] are on (LcdData0[1:0]=11 in RegLcdData0) and segments connected to pad_lcd_io[1] are off (LcdData0[5:4]=00 in RegLcdData0). The figure shows on the left side the waveforms on the circuit pins, in the middle the segment status and on the right the voltage on the segment (difference between the segment pin pad_lcd_io and common signal pad_lcd_com0 or pad_lcd_com1). © Semtech 2006 www.semtech.com 19-10 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 19.4.3 1:3 Multiplex scheme With 1:3 multiplexing, each pad_lcd_io pin can drive three segments of the display. In this mode, up to 93 segments of the display can be driven by the pad_lcd_io[30:0] pins. The common nodes are to be connected to the pad_lcd_com0, pad_lcd_com1 and pad_lcd_io[31]. These connections are shown in Figure 19-6. pad_vgen_vb pad_lcd_vr1 pad_lcd_vr2 pad_vgen_v3 pad_vgen_v2 1.2V voltage reference pad_vgen_v1 C aux Clcd1 VGEN Clcd2 Clcd3 pad_vgen_va pad_lcd_io[30:0] pad_lcd_io[31] pad_lcd_com1 pad_lcd_com0 SEGMENTS LCD DISPLAY COM2 COM1 COM0 Figure 19-6: LCD with 1:3 multiplexing and on-chip 1/3 bias voltage generation. Lots of different configurations for the voltage generation can be used. Figure 19-6 shows the connections and external elements that are required if the display is to be driven from 3.6V independently from the circuit supply voltage. Other possible configurations are given in section 19.5.4. The recommended value for the external capacitors is 470nF. To operate the driver in this configuration, the registers should be loaded with the values in Table 19-15. Register RegVgenCfg0 RegLcdOn RegLcdSe RegLcdClkFrame Contents[7:0] xx110001 xxxxx010 11111111 110xxx10 Table 19-15. Register contents for 1:3 mux example © Semtech 2006 www.semtech.com 19-11 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver V3 pad_lcd_com0 V2 V1 V0 pad_lcd_com0 - pad_lcd_io[0] (on) V3 V3 V2 V2 pad_lcd_com1 V1 V1 V0 V0 -V1 pad_lcd_io[31] V3 -V2 V2 -V3 V1 pad_lcd_com1 - pad_lcd_io[1] (off) V0 V3 V2 V1 V3 pad_lcd_io[0] V0 V2 -V1 V1 -V2 V0 -V3 V3 pad_lcd_io[1] pad_lcd_io[31] - pad_lcd_io[2] (on) V2 V3 V1 V2 V0 V1 V0 V3 pad_lcd_io[2] -V1 V2 -V2 V1 -V3 V0 Figure 19-7. 1:3 MUX mode waveforms The voltage generator has to be configured for the use of the internal reference (VgenRefEn=1 in RegVgenCfg0). The generator has to run (VgenOff=0 and VgenStdb=0 in RegVgenCfg0). It has to generate 1/3 bias (VgenMode=0 in RegVgenCfg0) and the generator output impedance is set to minimum (VgenClkSel=11 in RegVgenCfg0). The LCD driver is set to 1:3 multiplexing (LcdMux=10 in RegLcdOn) and the waveform generator is enabled (LcdSleep=0 in RegLcdOn). All pads are set into the LCD driver mode by setting all bits of register RegLcdSe to 1. To select a frame rate of about 50Hz, set the bits LcdFreq=10 and LcdDivFreq=110 in RegLcdClkFrame. Note that the precision of the frame frequency depends on the selected clock source (see clock block documentation). The 93 segments are on or off depending on the bits set in the registers RegLcdDataN. Only the bits 0,1,2 and 4,5,6 of these registers are used in 1:3 multiplexing. The bits 3 and 7 are not important. Figure 19-7 shows the generated waveforms for nine segments. The segments connected to pad_lcd_io[0] are all on (LcdData0[2:0]=111 in RegLcdData0), the segments connected to pad_lcd_io[1] are all off (LcdData0[6:4]=000 in RegLcdData0) and the segments connected to pad_lcd_io[2] are partially on, partially off (LcdData1[2:0]=101 in RegLcdData1). The © Semtech 2006 www.semtech.com 19-12 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver figure shows on the left side the waveforms on the circuit pins, in the middle the segment status and on the right the voltage on some segments (difference between the segment pin pad_lcd_io and common signals: pad_lcd_com0, pad_lcd_com1 or pad_lcd_io[31]). 19.4.4 1:4 Multiplex scheme pad_vgen_vb pad_lcd_vr1 pad_lcd_vr2 pad_vgen_v3 pad_vgen_v2 1.2V voltage reference pad_vgen_v1 Caux Clcd1 VGEN Clcd2 Clcd3 pad_vgen_va pad_lcd_io[29:0] pad_lcd_io[30] pad_lcd_io[31] pad_lcd_com1 pad_lcd_com0 LCD DISPLAY SEGMENTS COM3 COM2 COM1 COM0 Figure 19-8: LCD with 1:4 multiplexing and on-chip 1/3 bias voltage generation. With 1:4 multiplexing, each pin pad_lcd_io can drive four segments of the display. In this mode, up to 120 segments of the display can be driven by the pad_lcd_io[29:0] pins. The common nodes are to be connected to the pad_lcd_com0, pad_lcd_com1, pad_lcd_io[31] and pad_lcd_io[30]. These connections are shown in Figure 19-8. Lots of different configurations for the voltage generation can be used. Figure 19-8 shows the connections and external elements that are required if the display is to be driven from V3=3.6V, V2=2.4V and V1=1.2V independently from the circuit supply voltage. Other possible configurations are given in section 19.5.4. The recommended value for the external capacitors is 470nF. © Semtech 2006 www.semtech.com 19-13 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver V3 p a d _ lc d _ c o m 0 V2 V1 V0 p a d _ lc d _ c o m 0 - p a d _ lc d _ io [ 0 ] (o n ) V3 V3 V2 p a d _ lc d _ c o m 1 V2 V1 V1 V0 V0 -V1 -V2 V3 p a d _ lc d _ io [3 1 ] -V3 V2 V1 p a d _ lc d _ c o m 1 - p a d _ lc d _ io [ 1 ] (o f f) V0 V3 V2 V3 p a d _ lc d _ io [3 0 ] V1 V2 V0 V1 -V1 V0 -V2 -V3 p a d _ lc d _ io [3 0 ] - p a d _ lc d _ io [2 ] ( o n ) V3 p a d _ lc d _ io [0 ] V3 V2 V2 V1 V1 V0 V0 -V1 V3 -V2 p a d _ lc d _ io [1 ] V2 -V3 V1 V0 V3 p a d _ lc d _ io [2 ] V2 V1 V0 Fr a me f r e q u e n cy Figure 19-9. 1:4 MUX mode waveforms To operate the driver in this configuration, the registers should be loaded with the values in Table 19-16. The voltage generator has to be configured for the use of the internal reference (VgenRefEn=1 in RegVgenCfg0). The generator has to run (VgenOff=0 and VgenStdb=0 in RegVgenCfg0). It has to generate 1/3 bias (VgenMode=0 in RegVgenCfg0) and the generator output impedance is set to minimum (VgenClkSel=11 in RegVgenCfg0). Register RegVgenCfg0 RegLcdOn RegLcdSe RegLcdClkFrame Contents[7:0] xx110001 xxxxx011 11111111 100xxx10 Table 19-16. Register contents for 1:4 mux example The LCD driver is set to 1:4 multiplexing (LcdMux=11 in RegLcdOn) and the waveform generator is enabled (LcdSleep=0 in RegLcdOn). All pads are set into the LCD driver mode by setting all bits of register RegLcdSe to © Semtech 2006 www.semtech.com 19-14 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 1. To select a frame rate of about 50Hz, set the bits LcdFreq=10 and LcdDivFreq=100 in RegLcdClkFrame. Note that the precision of the frame frequency depends on the selected clock source (see clock block documentation). The 120 segments are on or off depending on the bits set in the registers RegLcdDataN. All bits of these registers are used in 1:4 multiplexing. Figure 19-9 shows the generated waveforms for twelve segments. The segments connected to pad_lcd_io[0] are all on (LcdData0[3:0]=1111 in RegLcdData0), the segments connected to pad_lcd_io[1] are all off (LcdData0[7:4]=0000 in RegLcdData0) and the segments connected to pad_lcd_io[2] are partially on, partially off (LcdData1[2:0]=1101 in RegLcdData1). The figure shows on the left side the waveforms on the circuit pins, in the middle the segment status and on the right the voltage on some segments (difference between the pad_lcd_io pin and common signal: pad_lcd_com0, pad_lcd_com1, pad_lcd_io[31] or pad_lcd_io[30]). 19.5 Advanced LCD features 19.5.1 Register usage To set the peripheral in LCD mode, all bits in RegLcdSe have to be set to 1. The status of the segments has to be written to the registers RegLcdDataN (with 0<N<15). Each register drives the segments connected to the pins pad_lcd_io[2N] (bits 0 to 3) and pad_lcd_io[2N+1] (bits 4 to 7). Depending on the selected multiplexing (LcdMux in RegLcdOn), not all bits in the data registers RegLcdDataN are used as shown in Table 19-17. The unused locations will always read back ‘0’. When reading back these registers, it will give the actual value on the display. When writing to these registers, the modifications will be taken into account only at the start of a new frame. Note that in the 1:3 multiplexing, the pad_lcd_io[31] pin is used as the third common signal and the data LcdData15[7:4] are not used. In the 1:4 multiplexing, the pad_lcd_io[31:30] are used as common signals and the data LcdData15[7:0] are not used. RegLcdDataN[x] 3, 7 2, 6 1, 5 0, 4 DD x x x used 1:2 x x used used 1:3 x used used used 1:4 used used used used Table 19-17. Useful data. In the LCD mode, all pads have LCD driver capabilities. The parallel port configuration registers (RegPLcdPullupN, RegPLcdDirN and RegPLcdOutN ) are unused. Writing to these registers will have no influence on the peripheral activity. These registers and RegPLcdInN will read back ‘0’. 19.5.2 Sleep mode or blinking mode When putting the LCD driver in sleep mode (LcdSleep = 1 in RegLcdOn), the peripheral will stop the waveform generation at the end of a frame. Entering sleep mode does not modify the LCD set-up but during sleep mode, the user is still allowed to write data into the RegLcdDataN registers. Figure 19-10 shows an example of the sleep mode synchronisation. The sleep signal can be used to blink the display. © Semtech 2006 www.semtech.com 19-15 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver V3 V2 V1 V0 V3 V2 V1 V0 LcdSleep Frame 1 Frame 0 Sleep Frame 2 Figure 19-10. Sleep mode synchronisation 19.5.3 LCD frame frequency selection 512 Hz 1024 Hz 2048 Hz 4096 Hz 00 01 10 Frame rate frequency 3-bit Programmable divider Internal State generator 11 Lcdfreq(1:0) Lcddivfreq(2:0) LcdMux(1:0) Figure 19-11: Frame Rate generation The LCD driver frame frequency depends on three parameters: the selected clock source frequency, the selected division factor and the multiplexing scheme. As shown in Figure 19-11, four different clock source frequencies can be selected by setting the bits LcdFreq in RegLcdClkFrame. This frequency is further divided by a factor between 1 and 8 depending on the bits LcdDivFreq in RegLcdClkFrame. It further depends on the selected multiplexing scheme (bits LcdMux in RegLcdOn). The selections are defined in Table 19-18 to Table 19-20. The frame rate FR is then given by: FR = f LCD 2 ⋅ div ⋅ mux Finally, Table 19-21 shows the minimal and maximal frame rate for each multiplexing. The source frequency should be chosen as low as possible to reduce the power consumption. Note that the precision of the frame rate depends on the precision of the selected clock source (see clock block documentation). © Semtech 2006 www.semtech.com 19-16 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver LcdFreq f LCD [Hz] 00 01 10 11 512 1048 2048 4096 Table 19-18. LCD driver source frequency selection LcdDivFreq 000 001 010 011 100 101 110 111 div 1 2 3 4 5 6 7 8 Table 19-19. LCD driver frequency division LcdMux 00 01 10 11 multiplexing scheme direct drive 1:2 1:3 1:4 mux 1 2 3 4 Table 19-20. LCD multiplexing LcdMux 00 00 01 01 10 10 11 11 LcdFreq 00 11 00 11 00 11 00 11 LcdDivFreq 111 000 111 000 111 000 111 000 FR [Hz] 32 2048 16 1024 10.6 682.7 8 512 Table 19-21. Minimal and maximal Frame Rate 19.5.4 Voltage generation The different voltages used in the LCD driver can be generated in several ways. Depending on the selected multiplexing scheme, the LCD display driver needs 2, 3 or 4 different voltages. The reference voltage can be derived from an internal 1.2V band gap reference or from an externally supplied voltage. The circuit also contains a voltage multiplier/divider that can be configured to generate a 1/2 bias or a 1/3 bias. This multiplier/divider needs external capacitors. Instead of the internal multiplier/divider, an external resistive ladder can be used. The use of the internal circuitry will in most applications give the lowest power solution. The voltages V0, V1 and V2 are connected internally in the circuit. The voltage V3 is not connected internally but externally by short-circuiting the pins pad_lcd_vr1 and pad_lcd_vr2 to pad_vgen_v3. The voltages used in the LCD driver for the different multiplexing schemes are given in Table 19-22. © Semtech 2006 www.semtech.com 19-17 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver multpilexing direct drive 1:2 1:3 1:4 V0 used used used used V1 used used V2 used used used V3 used used used used Table 19-22. V0, V1, V2, V3 usage in the LCD driver 19.5.4.1 Generating LCD voltage with the internal multiplier/divider. To generate LCD voltage with the internal multiplier/divider, some external capacitors have to be connected to the circuit. The value of the capacitor C aux can be calculated depending on the output impedance that is required on the V3 voltage: C aux ≅ with 6 f vgen ⋅ Z outV 3 f vgen the operating frequency of the multiplier/divider (set by the bits VgenClkSel in RegVgenCfg0, see Table 19-23) and Z outV 3 the required output impedance of V3. The equation is valid for C aux <5µF. Note that the operating frequency depends on the selected clock source (see clock block documen-tation). For a capacitor C aux of 470nF and a frequency of 1024Hz (default value), an output impedance of 12kΩ is obtained. The capacitors CLCD1, CLCD2 and CLCD3 can be chosen equal to C aux . VgenClkSel f vgen (Hz) 00 01 10 11 256 512 1024 2048 Table 19-23. multplier/divider operating frequency To enable the voltage multiplier/divider, the bits VgenOff and VgenStdb in RegVgenCfg0 are set to 0. The difference between the two bits is that VgenOff stops the generator and forces the nodes pad_vgen_v1, pad_vgen_v2, pad_vgen_v3, pad_vgen_va and pad_vgen_vb to predefined values and therefor discharges the capacitors CLCD1, CLCD2 and CLCD3. The bit VgenStdb only stops the operating clock without changing the voltage on CLCD1, CLCD2 and CLCD3. The capacitors will be discharged by leakage and by the switching of the LCD if the bit LcdSleep is not set. The multiplier/divider can generate 1/2 bias (VgenMode=1 in RegVgenCfg0) or 1/3 bias (VgenMode=0 in RegVgenCfg0). Finally, the multiplier/divider can use the internal 1.2V bandgap reference (VgenRefEn=1 in RegVgenCfg0) or an external reference (VgenRefEn=0 in RegVgenCfg0). The internal voltage reference can not be used in case the circuit voltage supply is below 1.5V. Figure 19-12 shows the external connections to be made in 1/3 bias mode. Part (a) of the figure shows the use of the internal voltage reference. In this case, V1=1.2V, V2=2.4V and V3=3.6V. Part (b) of the figure shows the use of an external voltage reference. The external reference can be connected to one of the pins pad_vgen_v1, pad_vgen_v2, pad_vgen_v3. Table 19-24 shows the voltage generated on V1, V2, V3 depending on the connection of the external reference voltage Vext. The external voltage may or may not be identical to the circuit © Semtech 2006 www.semtech.com 19-18 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver supply voltage VBAT. The voltage on V3 should not exceed 5.5V. The voltage on pad_vgen_v1 should not exceed VBAT. VgenOff = 0 VgenRefEn = 1 VgenMode = 0 VgenOff = 0 VgenRefEn = 0 VgenMode = 0 LCD driv er LCD driv er pad_v gen_v b pad_v gen_v b pad_v gen_v 3 pad_v gen_v 3 Clcd3 Clcd3 pad_v gen_v 2 pad_v gen_v 2 Clcd2 pad_v gen_v 1 1.2V v oltage ref erence Clcd2 Caux Clcd1 Caux Clcd1 VGEN pad_v gen_v a VGEN 1.2V v oltage ref erence pad_v gen_v 1 pad_v gen_v a (b) (a) v oltage Vext Figure 19-12: Generation of LCD voltage with external capacitors for 1/3 bias mode. Vext connection V1 (V) V2 (V) V3 (V) pad_vgen_v1 pad_vgen_v2 pad_vgen_v3 Vext (1/2)⋅Vext (1/3) ⋅Vext 2⋅Vext Vext (2/3) ⋅Vext 3⋅Vext (3/2)⋅Vext Vext Table 19-24. V1, V2 and V3 as a function of the external reference connection in 1/3 bias mode Figure 19-13 shows the external connections to be made in 1/2 bias mode. Part (a) of the figure shows the use of the internal voltage reference. In this case, V1=V2=1.2V and V3=2.4V. Part (b) of the figure shows the use of an external voltage reference. The external reference can be connected to one of the pins pad_vgen_v1 or pad_vgen_v3. Table 19-25 shows the voltage generated on V1 and V3 depending on the connection of the external reference voltage Vext. The external voltage may or may not be identical to the circuit supply voltage VBAT. The voltage on V3 should not exceed 5.5V. Vext connection V1 (V) V3 (V) pad_vgen_v1 pad_vgen_v3 Vext (1/2) ⋅Vext 2⋅Vext Vext Table 19-25. V1 and V3 as a function of the external reference connection in 1/2 bias mode © Semtech 2006 www.semtech.com 19-19 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver VgenOff = 0 VgenRefEn = 1 VgenMode = 1 VgenOff = 0 VgenRefEn = 0 VgenMode = 1 LCD driv er LCD driv er pad_v gen_v b pad_v gen_v b pad_v gen_v 3 pad_v gen_v 3 Clcd3 Clcd3 pad_v gen_v 2 pad_v gen_v 2 pad_v gen_v 1 Caux pad_v gen_v 1 Caux 1.2V v oltage ref erence 1.2V v oltage ref erence Clcd1 Clcd1 pad_v gen_v a VGEN VGEN pad_gen_v a (b) (a) v oltage Vext Figure 19-13: Generation of LCD voltage with external capacitors for 1/2 bias mode. Finally, in direct drive, there is no need for intermediate voltages. The configurations of Figure 19-12(a) or Figure 19-13(a) can still be used (depending on the required voltage: 2.4V or 3.6V) but the configurations in Figure 19-12(b) or Figure 19-13(b) can be replaced by the simple schematic of Figure 19-14. The external voltage Vext may or may not be identical to the circuit supply voltage VBAT. VgenOff = 1 VgenRefEn = 0 LCD driver pad_gen_v b pad_gen_v 3 pad_gen_v 2 pad_v gen_v 1 1.2V voltage reference voltage Vext pad_gen_v a VGEN Figure 19-14. Generation of LCD voltage for direct drive using external voltage. 19.5.4.2 Generating LCD voltages with an external resistor ladder To generate the LCD voltages with an external R-ladder (Figure 19-15), the internal voltage reference and multiplier/divider block are not used (VgenOff=1 and VgenRefEn=0 in RegVgenCfg0). All other bits in the register RegVgenCfg0 are unused. All resistor values are equal and sufficiently small in order to have small LCD voltage impedance. For direct drive, no resistors are required as shown in Figure 19-14. The voltage Vext can be connected to the circuit supply voltage VBAT or any other external voltage. © Semtech 2006 www.semtech.com 19-20 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver VgenOff = 1 VgenRefEn = 0 LCD driver VgenOff = 1 VgenRefEn = 0 LCD driver Vext Vext pad_gen_v b pad_gen_v b pad_v gen_v 3 pad_v gen_v 3 pad_v gen_v 2 R3 pad_v gen_v 1 R2 1.2V voltage reference R3 pad_v gen_v 2 pad_v gen_v 1 1.2V voltage reference R1 VGEN R1 VGEN pad_v gen_v a pad_v gen_v a (b) (a) Figure 19-15. Generation of LCD voltages with external resistors ladder in 1/3 (a) and 1/2 (b) bias mode. 19.6 Parallel I/O port capabilities 19.6.1 Parallel port function All pins pad_lcd_io[31:0] can be used as a general input/output digital port. The peripheral is set to this mode by writing all bits of the register RegLcdSe to ‘0’. Figure 19-16 shows the structure for one pin and the registers used in this mode. Since the peripheral has 32 pins, 4 of each of the registers in Figure 19-16 exist. They are labelled with the suffix 0 to 3. The mapping of the register bits to the I/O pins is given in Table 19-26. As an example, the bit PLcdPullup2[6] in RegPLcdPullup2 controls the pull up resistor of the pin pad_lcd_io[22]. suffix of the register 0≤N≤3 N bit in the register 0≤n≤7 n pin pad_lcd_io[8N+n] Table 19-26. Register bit to pin mapping The direction of each pin pad_lcd_io[31:0] (input or output) can be individually set by using the RegPLcdDirN registers. If PLcdDirN[n] = 1, the output buffer of the corresponding pad_lcd_io[8N+n] is enabled. Output mode: The data to be output is stored in RegPLcdOutN. In output mode, the pull-up resistors should be switched off by writing ‘0’ to the registers RegPLcdPullupN. If not, current will flow through the pull-up resistors when the output is forced to ‘0’. © Semtech 2006 www.semtech.com 19-21 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver PORT R RegPlcdPullup RegPlcdOut RegPlcdDir RegPlcdIn Figure 19-16. Structure of the LCD driver used as parallel port Input mode: The status of the pins pad_lcd_io is available in RegPLcdInN (read only). Reading is always direct, there is no digital debounce function. In case of noisy input signals, a software debouncer or an external filter must be realised. The pull-up resistors are individually controllable for each pin by setting the corresponding bit in the registers RegPLcdPullN (1=active, 0=inactive). 19.6.2 Parallel Port Voltage When the peripheral is used as a parallel port, the internal voltage generator is not required (VgenOff=1 and VgenRefEn=0 in RegVgenCfg0). For normal operation as a standard parallel port, the circuit is connected as shown in Figure 19-17. In this case, the logical ‘1’ corresponds to the voltage VBAT. The parallel port can also be driven from another voltage than VBAT. This allows the circuit to be interfaced to other circuits that have a different supply voltage domain VDD1 that can be either lower or higher than VBAT without adding any extra hardware as shown in Figure 19-18. The parallel port can be split in several voltage domains. The pins pad_lcd_io[31:12] are supplied from the pad_vgen_v3 pin. The pins pad_lcd_io[11:4] are supplied from the pin pad_lcd_vr1 and the pins pad_lcd_io[3:0] are supplied from the pin pad_lcd_vr2. Each of these can be supplied with a different voltage as shown in Figure 19-19. As an example, VBAT could be supplied at 2.7V, VDD1 at 2.2V, VDD2 at 3.3V and VDD3 at 5V. The only limitation is that VDD1>VREG, VDD2>VREG, VDD3>VREG. © Semtech 2006 www.semtech.com 19-22 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver VgenOff = 1 VgenRefEn = 0 V1 V2 V3 VBAT pad_lcd_vr1 pad_lcd_vr2 pad_vgen_vb pad_vgen_v3 pad_vgen_v2 pad_vgen_v1 1.2V voltage reference pad_vgen_va VGEN VSS Figure 19-17. Parallel port voltage connection VgenOff = 1 VgenRefEn = 0 VDD1 domain VBAT pad_lcd_io[31:0] I/O[31:0] VDD1 pad_lcd_vr1 pad_lcd_vr2 VDD1 VSS pad_vgen_vb pad_vgen_v3 pad_vgen_v2 pad_vgen_v1 1.2V voltage reference VGEN pad_vgen_va VSS Figure 19-18. Parallel port voltage connection with I/O levelshifting © Semtech 2006 www.semtech.com 19-23 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver VBAT vbat VDD1 VBAT domain pad_lcd_vr2 VDD1 I/Os pad_lcd_io[3:0] VDD1 domain vgen_off = 1 vgen_ref_en = 0 VDD2 VDD2 I/Os pad_lcd_vr1 pad_lcd_io[11:4] VDD2 domain VDD3 pad_vgen_vb pad_vgen_v3 pad_lcd_io[31:12] VDD3 I/Os VDD3 domain 1.2V voltage reference pad_vgen_v2 pad_vgen_v1 VGEN pad_vgen_va vss Figure 19-19. Multi voltage parallel port interface 19.7 Partial LCD Driver / Partial Parallel I/O Port It is perfectly possible to combine different modes described in the previous sections on different pins. Using the bits LcdSe in the register RegLcdSe, the pins pad_lcd_io[31:0] can be set as parallel port or as LCD drivers. Each bit sets the mode for 4 pins (see Table 19-4). When combining the LCD and digital port functions, care has to be taken with the voltage generation. The logical ‘1’ of the digital I/O is driven from the V3 rail and needs to be low impedance. The V3 of the pins used for digital I/O can not be supplied by the on-circuit voltage multiplier. The next sections show different possibilities. 19.7.1 Single low impedance voltage for V3 of LCD and digital I/O If V3 in the LCD driver is supplied from a low impedance voltage supply (as e.g. in Figure 19-12(b) or Figure 19-13(b) with connection of Vext to pad_vgen_v3, or in Figure 19-14 and Figure 19-15), the partitioning of the pins pad_lcd_io[31:0] between the LCD driver function and the digital parallel port function can be chosen freely by setting the bits LcdSe. Vext may be equal to the circuit supply voltage VBAT but this is not required. The digital I/O pins use the voltages V0=VSS and V3=Vext for the logical ‘0’ and logical ‘1’ respectively. © Semtech 2006 www.semtech.com 19-24 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver vdd vbat pad_vgen_vb pad_lcd_vr1 pad_lcd_vr2 External circuit pad_vgen_v3 pad_vgen_v2 1.2V voltage reference pad_vgen_v1 Caux Clcd1 Clcd3 vss vss VGEN pad_vgen_va pad_lcd_io[15:0] IO_1[15:0] IO_2[7:0] pad_lcd_io[31:24] pad_lcd_io[23:16] pad_lcd_com1 pad_lcd_com0 SEGMENTS[7:0] LCD DISPLAY COM1 COM0 Figure 19-20: Sharing LCD (1:2 mux) and digital I/O (low impedance V3) Figure 19-20 and Table 19-27 show a possible example for such a configuration. In this case, the LCD driver voltage V3=VBAT and V1=VBAT/2 (1/2 bias mode, VgenMode=1, VgenRefEn=0 and VgenOff=0 in RegVgenCfg0). The LCD is in 1:2 multplexing (LcdMux=01 in RegLcdOn) and LcdSe19=LcdSe23=1 while all other bits in RegLcdSe are 0. Register RegVgenCfg0 RegLcdOn RegLcdSe Contents[7:0] xx110100 xxxxx001 11110011 Table 19-27. Register contents for configuration of Figure 19-20. © Semtech 2006 www.semtech.com 19-25 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 19.7.2 Different Voltages for V3 of LCD and Digital I/O When configured as a digital I/O, the logical ‘1’ on the pins pad_lcd_io[31:0] is identical to V3. For a part of the pad_lcd_io[31:0] pins, the V3 connection is not made inside the circuit but has to be done externally by connecting the pins pad_vgen_v3, pad_lcd_vr1 and pad_lcd_vr2 together. This feature allows for the use of different voltages on V3 for the LCD display and for I/O parallel pins. Table 19-28 shows the partitioning of the pins. The voltages that can be applied on pad_vgen_v3, pad_lcd_vr1, pad_lcd_vr2 and VBAT are completely independent from each other. pin pad_lcd_com0 pad_lcd_com1 pad_lcd_io[31:12] pad_lcd_io[11:4] pad_lcd_io[3:0] V3 connection pad_vgen_v3 pad_lcd_vr1 pad_lcd_vr2 Table 19-28. V3 connection of the different pad_lcd_io pins vbat vbat vdd1 pad_lcd_vr2 vdd pad_lcd_io[3:0] io[3:0] EXTERNAL CIRCUIT vss vdd1 voltage domain vdd2 pad_lcd_vr1 vdd io[7:0] pad_lcd_io[11:4] EXTERNAL CIRCUIT vss vdd2 voltage domain pad_vgen_vb pad_vgen_v3 pad_vgen_v2 1.2V voltage reference pad_vgen_v1 Caux Clcd1 VGEN Clcd2 Clcd3 vss pad_vgen_va pad_lcd_io[29:12] pad_lcd_io[30] pad_lcd_io[31] pad_lcd_com1 pad_lcd_com0 SEGMENTS[17:0] LCD DISPLAY COM3 COM2 COM1 COM0 Figure 19-21. Sharing LCD driver (1:4 mux) and digital I/O with V3 ≠ logic ‘1’ © Semtech 2006 www.semtech.com 19-26 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver As can be seen from Table 19-28, the voltage V3 on the pins pad_lcd_io[31:12] can not be dissociated from the voltage V3 on the pins pad_lcd_com0, pad_lcd_com1 and the internal voltage multiplier/divider. It means that, if V3 is not a low impedance external voltage as in the previous section, they can be used for the LCD driver only and not for digital I/O. Figure 19-21 and Table 19-29 show an example. In this case, the pins pad_lcd_io[29:12] are used to drive a display with 1:4 multiplexing (LcdSe15=LcdSe19=LcdSe23=Lcd27=Lcd31=1 in RegLcdSe and LcdMux=11 in RegLcdOn). In 1:4 multiplexing, the lines pad_lcd_io[31:30] are used for COM2 and COM3. The voltage V3 for the display is generated by the internal voltage multiplier/divider using the internal reference (VgenOff=0, VgenRefEn=1, VgenMode=0 in RegVgenCfg0). The segment status is set by using the RegLcdDataN registers with 6≤N≤15. Writing in the registers with 0≤N≤5 will have no effect. The pins pad_lcd_io[11:0] are used as digital I/O (LcdSe3=LcdSe7=LcdSe11=0 in RegLcdSe). The control of the digital I/O is done using the registers RegPLcdInN, RegPLcdOutN, RegPLcdDirN and RegPLcdPullupN with 0≤N≤1. Writing in the registers with 2≤N≤3 will have no effect. The pins pad_lcd_io[11:4] and pad_lcd_io[3:0] can further be split into two different voltage domains. The voltage domains VDD1, VDD2, V3 and VBAT are independent. The only limitation is that VDD1>VREG and VDD2>VREG. In the example V3=3.6V, VBAT could be at 2.7V, VDD1 at 2.4V and VDD2 at 5V. Register RegVgenCfg0 RegLcdOn RegLcdSe Contents[7:0] xx110001 xxxxx011 00011111 Table 19-29. Register contents for configuration of Figure 19-21. 19.8 Specifications 19.8.1 pad_lcd_io used in LCD mode Specification V1 V2, V3 trise-fall Min 1.1 1.1 Typ Max VBAT 5.5 25 Unit V V µs Description Comments Rise/Fall time (LCD mode) (1) (2) (3) (1) rise or fall time from 10% to 90% of the output signal (2) Cload=5000pF (3) V1=V2/2=V3/3=1.1V (1/3 bias) or V1=V2=V3/2=1.1V (1/2 bias) 19.8.2 pad_lcd_io used in digital I/O mode Specification pad_lcd_vr1 pad_lcd_vr2 pad_vgen_v3 R_pullud trise-fall IOD Min VREG VREG VREG 35 Typ 1 8 Max 5.5 5.5 5.5 100 Unit V V V kΩ µs mA Description pad supply voltage pad supply voltage pad supply voltage Pull up/down resistance Rise/Fall time Output current drive Comments (1) (2) (3) (1) rise or fall time from 10% to 90% of the output signal (2) with Cload=5nF, pad_vgen_v3=pad_lcd_vr1=pad_lcd_vr2=2.4V (3) pad_vgen_v3=pad_lcd_vr1=pad_lcd_vr2=4.5V, voltage on pad_lcd_io=0.4V for sink current and 4.1V for source current. © Semtech 2006 www.semtech.com 19-27 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 19.8.3 Voltage Reference Specification Vref ∆Vref/∆T Power supply VBAT I_load on V1 Zout on V1 19.8.4 (5) Typ 1.17 0.2 1.5 Max 1.34 5.5 2 1000 Unit V mV/°C V mA Ω Comment @20°C, VBAT>2.4V LCD multiplier/divider Specification Settling time to 90% Zout,v2 on V2 Zout,v3 on V3 pad_vgen_v3 (1) (2) (3) (4) Min 1.0 Min Typ 3 (1) 7 (1) 6 (2) 14 (2) Max 30 25 (3) 60 (3) 5.5 Unit ms kΩ kΩ V Comments (2), (4), (5) (4), (5) (4), (5) f(vgen_clk) = 2 kHz. f(vgen_clk) = 1 kHz. f(vgen_clk) = 0.25 kHz. 1/3 bias mode. with 0.47µF external capacitors. © Semtech 2006 www.semtech.com 19-28 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 20. 20.1 20.2 20.3 20.4 20.5 20.6 20.7 20.8 20.9 20.10 20.11 20.12 Counters/PWM FEATURES ..........................................................................................................................20-2 OVERVIEW ..........................................................................................................................20-2 REGISTER MAP ...................................................................................................................20-2 INTERRUPTS AND EVENTS MAP .............................................................................................20-4 BLOCK SCHEMATIC ..............................................................................................................20-4 GENERAL COUNTER REGISTERS OPERATION ..........................................................................20-5 CLOCK SELECTION ..............................................................................................................20-5 COUNTER MODE SELECTION .................................................................................................20-6 COUNTER / TIMER MODE ......................................................................................................20-7 PWM MODE .......................................................................................................................20-8 CAPTURE FUNCTION ..........................................................................................................20-10 SPECIFICATIONS ...............................................................................................................20-11 © Semtech 2006 www.semtech.com 20-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 20.1 Features - 4 x 8-bits timer/counter modules or 2 x 16-bits timers/counter modules Each with 4 possible clock sources Up/down counter modes Interrupt and event generation Capture function (internal or external source) Rising, falling or both edge of capture signal (except for xtal 32 kHz, only rising edge) PA[3:0] can be used as clock inputs (debounced or direct, frequency divided by 2 or not) 2 x 8 bits PWM or 2 x 16 bits PWM PWM resolution of 8, 10, 12, 14 or 16 bits Complex mode combinations are possible 20.2 Overview Counter A and Counter B are 8-bits counters and can be combined to form a 16-bit counter. Counter C and Counter D exhibit the same feature. The counters can also be used to generate two PWM outputs on PB[0] and PB[1]. In PWM mode one can generate PWM functions with 8, 10, 12, 14 or 16 bits wide counters. The counters A and B can be captured by events on an internal or an external signal. The capture can be performed on both 8-bit counters running individually on two different clock sources or on both counters chained to form a 16-bit counter. In any case, the same capture signal is used for both counters. When the counters A and B are not chained, they can be used in several configurations: A and B as counters, A and B as captured counters, A as PWM and B as counter, A as PWM and B as captured counter. When the counters C and D are not chained, they can be used either both as counters or counter C as PWM and counter D as counter. 20.3 Register map register name RegCntA RegCntB RegCntC RegCntD RegCntCtrlCk RegCntConfig1 RegCntConfig2 RegCntOn Table 20-1. Counter registers bit 7-0 7-0 RegCntA CounterA CounterA rw R W reset 00000000 nresetglobal 00000000 nresetglobal function 8-bits counter value 8-bits comparison value Table 20-2. RegCntA © Semtech 2006 www.semtech.com 20-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver bit 7-0 7-0 RegCntB CounterB CounterB rw R W reset 00000000 nresetglobal 00000000 nresetglobal function 8-bits counter value 8-bits comparison value Table 20-3. RegCntB Note: When writing to RegCntA or RegCntB, the processor writes the counter comparison values. When reading these locations, the processor reads back either the actual counter value or the last captured value if the capture mode is active. bit 7-0 7-0 RegCntC CounterC CounterC rw R W reset 00000000 nresetglobal 00000000 nresetglobal function 8-bits counter value 8-bits comparison value Table 20-4. RegCntC bit 7-0 7-0 RegCntD CounterD CounterD rw R W reset 00000000 nresetglobal 00000000 nresetglobal function 8-bits counter value 8-bits comparison value Table 20-5. RegCntD Note: When writing RegCntC or RegCntD, the processor writes the counter comparison values. When reading these locations, the processor reads back the actual counter value. bit 7-6 5-4 3-2 1-0 RegCntCtrlCk CntDCkSel(1:0) CntCCkSel(1:0) CntBCkSel(1:0) CntACkSel(1:0) rw R/W R/W R/W R/W reset 00 nresetglobal 00 nresetglobal 00 nresetglobal 00 nresetglobal function Counter d clock selection Counter c clock selection Counter b clock selection Counter a clock selection Table 20-6. RegCntCtrlCk bit 7 6 5 4 3 2 1 0 RegCntConfig1 CntDDownUp CntCDownUp CntBDownUp CntADownUp CascadeCD CascadeAB CntPWM1 CntPWM0 rw R/W R/W R/W R/W R/W R/W R/W R/W reset 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function Counter d up or down counting (0=down) Counter c up or down counting (0=down) Counter b up or down counting (0=down) Counter a up or down counting (0=down) Cascade counter c & d (1=cascade) Cascade counter a & b (1=cascade) Activate pwm1 on counter c or c+d (PB(1)) Activate pwm0 on counter a or a+b (PB(0)) Table 20-7. RegCntConfig1 bit 7-6 5-4 3-2 1-0 RegCntConfig2 CapSel(1:0) CapFunc(1:0) Pwm1Size(1:0) Pwm0Size(1:0) rw R/W R/W R/W R/W reset 00 nresetglobal 00 nresetglobal 00 nresetglobal 00 nresetglobal function Capture source selection Capture function Pwm1 size selection Pwm0 size selection Table 20-8. RegCntConfig2 © Semtech 2006 www.semtech.com 20-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver bit 7 6 5 4 3 2 1 0 RegCntOn rw R/W R/W R/W R/W R/W R/W R/W R/W CntDExtDiv CntCExtDiv CntBExtDiv CntAExtDiv CntDEnable CntCEnable CntBEnable CntAEnable reset 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function Divide PA(3) frequency by 2 (1=divide) Divide PA(2) frequency by 2 (1=divide) Divide PA(1) frequency by 2 (1=divide) Divide PA(0) frequency by 2 (1=divide) Enable counter d Enable counter c Enable counter b Enable counter a Table 20-9. RegCntOn 20.4 Interrupts and events map Interrupt source IrqA IrqB IrqC IrqD Default mapping in the interrupt manager RegIrqHigh(4) RegIrqLow(5) RegIrqHigh(3) RegIrqLow(4) Default mapping in the event manager RegEvn(7) RegEvn(3) RegEvn(6) RegEvn(2) Table 20-10. Default interrupt and event mapping. 20.5 Block schematic ck1k ck16k ck128 ckrcext/4 ckrcext RegCntA (write) RegCntA (read) Counter A PA(0) Capture RegCntB (write) RegCntB (read) Counter B PA(1) RegCntC (write) ck1k ck32k RegCntC (read) Counter C PA(2) RegCntD (write) RegCntD (read) Counter D PA(3) PB(0) PWM PB(1) Figure 20-1: Counters/timers block schematic © Semtech 2006 www.semtech.com 20-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 20.6 General counter registers operation Counters are enabled by CntAEnable, CntBEnable, CntCEnable, and CntDEnable in RegCntOn. To stop the counter X, CntXEnable must be reset. To start the counter X, CntXEnable must be set. When counters are cascaded, CntAEnable and CntCEnable also control respectively the counters B and D. WARNING: THERE SHOULD BE AT LEAST ONE CPU INSTRUCTION BETWEEN THE CONFIGURATION OF THE COUNTERS AND THE ENABLING. All counters have a corresponding 8-bit read/write register: RegCntA, RegCntB, RegCntC, and RegCntD. When read, these registers contain the counter value (or the captured counter value). When written, they modify the counter comparison values. It is possible to read any counter at any time, even when the counter is running. The value is guaranteed to be correct when the counter is running on an internal clock source. For a correct acquisition of the counter value when running on an external clock source, use one of the three following methods: 1) For slow operating counters (typically at least 8 times slower than the CPU clock), oversample the counter content and perform a majority operation on the consecutive read results to select the correct actual content of the counter. 2) Stop the concerned counter, perform the read operation and restart the counter. While stopped, the counter content is frozen and the counter does not take into account the clock edges delivered on the external pin. 3) Use the capture mechanism. When a value is written into the counter register while the counter is in counter mode, both the comparison value is updated and the counter value is modified. In upcount mode, the register value is reset to zero. In downcount mode, the comparison value is loaded into the counter. Due to the synchronization mechanism between the processor clock domain and the external clock source domain, this modification of the counter value can be postponed until the counter is enabled and that it receives it’s first valid clock edge. In the PWM mode or in the capture mode, the counter value is not modified by the write operation in the counter register. Changing to the counter mode, does not update the counter value (no reset in upcount, no load in downcount mode). 20.7 Clock selection The clock source for each counter can be individually selected by writing the appropriate value in the register RegCntCtrlCk. Clock source for CntXCkSel(1:0) 11 10 01 00 CounterA CounterB CounterC CounterD Ck128 CkRcExt/4 CkRcExt PA(0) PA(1) Ck1k Ck32k PA(2) PA(3) Table 20-11: Clock sources for counters A, B, C and D Table 20-11 gives the correspondence between the binary codes used for the configuration bits CntACkSel(1:0), CntBCkSel(1:0), CntCCkSel(1:0) or CntDCkSel(1:0) and the clock source selected respectively for the counters A, B, C or D. © Semtech 2006 www.semtech.com 20-5 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver The CkRcExt clock is the RC oscillator or external clock. The clocks below 32kHz can be derived from the RC oscillator, the external clock source or the crystal oscillator (see the documentation of the clock block). A separate external clock source can be delivered on PortA for each individula counter. The external clock sources can be debounced or not by properly setting the PortA configuration registers. Additionally, the external clock sources can be divided by two in the counter block, thus enabling higher external clock frequencies, by setting the CntXExtDiv bits in the RegCntOn register. Switching between an internal and an external clock source can only be performed while the counter is stopped. The enabling or disabling of the external clock frequency division can only be performed while the counter using this clock is stopped, or when this counter is running on an internal clock source. 20.8 Counter mode selection Each counter can work in one of the following modes: 1) Counter, downcount & upcount 2) Captured counter, downcount & upcount (only counters A&B) 3) PWM, downcount & upcount 4) Captured PWM, downcount and upcount The counters A and B or C and D can be cascaded or not. In cascaded mode, A and C are the LSB counters while B and D are the MSB counters. Table 20-12 shows the different operation modes of the counters A and B as a function of the mode control bits. For all counter modes, the source of the down or upcount selection is given (either the bit CntADownUp or the bit CntBDownUp). Also, the mapping of the interrupt sources IrqA and IrqB and the PWM output on PB(0) in these different modes is shown. CascadeAB CountPWM0 CapFunc(1:0) 0 0 00 1 0 00 0 1 00 1 1 00 0 0 1 0 0 1 1 1 1x or x1 1x or x1 1x or x1 1x or x1 Counter A mode Counter B mode IrqA source IrqB source PB(0) function Counter 8b Downup: A Counter 8b Downup: B Counter A Counter B PB(0) Counter 16b AB Downup: A PWM 8b Counter 8b Downup: A Downup: B PWM 10 – 16b AB Downup A Captured Captured counter 8b counter 8b Downup: A Downup: B Counter AB - PB(0) - Counter B PWM A - - PWM AB Capture A Capture B PB(0) Captured counter 16b AB Downup: A Capture AB Capture AB PB(0) Capture A Capture B PWM A Capture AB Capture AB PWM AB Captured Captured PWM 8b counter 8b Downup: A Downup: B Captured 10 – 16b PWM (captured value on 16b) Downup: A Table 20-12: Operating modes of the counters A and B Table 20-13 shows the different operation modes of the counters C and D as a function of the mode control bits. For all counter modes, the source of the down or upcount selection is given (either the bit CntCDownUp or the bit © Semtech 2006 www.semtech.com 20-6 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver CntDDownUp). The mapping of the interrupt sources IrqC and IrqD and the PWM output on PB(1) in these different modes is also shown. The switching between different modes must be done while the concerned counters are stopped. While switching capture mode on and off, unwanted interrupts can appear on the interrupt channels concerned by this mode change. CascadeCD CountPWM1 0 0 1 0 0 1 1 1 Counter C mode Counter D mode IrqC source IrqD source PB(1) function Counter 8b Counter 8b Downup: C Downup: D Counter 16b CD Downup: C PWM 8b Counter 8b Downup: C Downup: D PWM 10 – 16b CD Downup: C Counter C Counter CD - Counter D - PB(1) Counter D - PWM C - PB(1) PWM CD Table 20-13: Operating modes of the counters C and D 20.9 Counter / Timer mode The counters in counter / timer mode are generally used to generate interrupts after a predefined number of clock periods applied on the counter clock input. Each counter can be set individually either in upcount mode by setting CntXDownUp in the register RegCntConfig1 or in downcount mode by resetting this bit. Counters A and B can be cascaded to behave as a 16 bit counter by setting CascadeAB in the RegCntConfig1 register. Counters C and D can be cascaded by setting CascadeCD. When cascaded, the up/down count modes of the counters B and D are defined respectively by the up/down count modes set for the counters A and C. When in upcount mode, the counter will start incrementing from zero up to the target value which has been written in the corresponding RegCntX register(s). When the counter content is equal to the target value, an interrupt is generated at the next counter clock pulse and the counter is loaded again with the zero value (Figure 20-2). When in downcount mode, the counter will start decrementing from the initial load value which has been written in the corresponding RegCntX register(s) down to the zero value. Once the counter content is equal to zero, an interrupt is generated at the next counter clock pulse and the counter is loaded again with the load value (Figure 20-2). Be careful to select the counter mode (no capture, not PWM, specify cascaded or not and up or down counting mode) before writing any target or load value to the RegCntX register(s). This ensures that the counter will start from the correct initial value. When counters are cascaded, both counter registers must be written to ensure that both cascaded counters will start from the correct initial values. The stopping and consecutive starting of a counter in counter mode without a target or load value write operation in between can generate an interrupt if this counter has been stopped at the zero value (downcount) or at it’s target value (upcount). This interrupt is additional to the interrupt which has already been generated when the counter reached the zero or the target value. Due to the synchronization between the CPU clock and the counter clock source, it may take up to 1 CPU clock cycle before the configuration changes written in the RegCntConfigX or RegCntX registers are becoming effective. In order to get correct operation of the counters, there should be at least 1 software instruction between the modification of RegCntConfigX or RegCntX and the enabling of the counters. © Semtech 2006 www.semtech.com 20-7 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver dow n counting clock counter X RegcntX_r XX RegCntX_w XX 3 2 1 0 3 2 1 0 3 2 1 0 3 write RegCntX CntXDownUp IrqX CntXEnable up counting clock counter X RegCntX_r RegCntX_w XX XX 0 1 2 3 0 1 2 3 0 1 2 3 3 write RegCntX CntXDownUp IrqX CntXEnable Figure 20-2. Up and down count interrupt generation. 20.10 PWM mode The counters can generate PWM signals (Pulse Width Modulation) on the PortB outputs PB(0) and PB(1). The PWM mode is selected by setting CntPWM1 and CntPWM0 in the RegCntConfig1 register. See Table 20-12 and Table 20-13 for an exact description of how the setting of CntPWM1 and CntPWM0 affects the operating mode of the counters A, B, C and D according to the other configuration settings. When CntPWM0 is enabled, the PWMA or PWMAB output value overrides the value set in bit 0 of RegPBOut in the Port B peripheral. When CntPWM1 is enabled, the PWMC or PWMCD output value overrides the value set in bit 1 of RegPBOut. The corresponding ports (0 and/or 1) of Port B must be set in digital mode and as output and either open drain or not and pull up or not through a proper setting of the control registers of the Port B. Counters in PWM mode count down or up, according to the CntXDownUp bit setting. No interrupts and events are generated by the counters that are in PWM mode. Counters do count circularly: they restart at zero or at the maximal value (either 0xFF when not cascaded or 0xFFFF when cascaded) when respectively an overflow or an underflow condition occurs in the counting. © Semtech 2006 www.semtech.com 20-8 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver The internal PWM signals are low as long as the counter contents are higher than the PWM code values written in the RegCntX registers. They are high when the counter contents are smaller or equal to these PWM code values. In order to have glitch free outputs, the PWM outputs on PB(0) and PB(1) are sampled versions of these internal PWM signals, therefore delayed by one counter clock cycle. The PWM resolution is always 8 bits when the counters used for the PWM signal generation are not cascaded. PWM0Size(1:0) and PWM1Size(1:0) in the RegCntConfig2 register are used to set the PWM resolution for the counters A and B or C and D respectively when they are in cascaded mode. The different possible resolutions in cascaded mode are shown in Table 20-14. Choosing a 16 bit PWM code which is higher than the maximum value that can be represented by the number of bits chosen for the resolution results in a PWM output which is always tied to 1. PwmXsize(1:0) 11 10 01 00 Resolution 16 bits 14 bits 12 bits 10 bits Table 20-14: Resolution selection in cascaded PWM mode Small PWM code Tlsmall Thsmall Large PWM code Tllarge Thlarge Tper Figure 20-3: PWM modulation examples The period of the PWM signal is given by the formula: Tper = 2 resolution f ckcnt The duty cycle ratio DCR of the PWM signal is defined as: DCR = DCR can be selected between 100 2 resolution Th Tper % and 100 %. DCR in % in function of the RegCntX content(s) is given by the relation: ⎛ 100(1 + RegCntX ) ⎞ ,100 ⎟ DCR = MIN ⎜ resolution 2 ⎝ ⎠ © Semtech 2006 www.semtech.com 20-9 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 20.11 Capture function The 16-bit capture register is provided to facilitate frequency measurements. It provides a safe reading mechanism for the counters A and B when they are running. When the capture function is active, the processor does not read anymore the counters A and B directly, but instead reads shadow registers located in the capture block. An interrupt is generated after a capture condition has been met when the shadow register content is updated. The capture condition is user defined by selecting either internal capture signal sources derived from the prescaler or from the external PA(2) or PA(3) ports. Both counters use the same capture condition. When the capture function is active, the A and B counters can either upcount or downcount. They do not count circularly: they restart at zero or at the maximal value (either 0xFF when not cascaded or 0xFFFF when cascaded) when respectively an overflow or an underflow condition occurs in the counting. The capture function is also active on the counters when used to generate PWM signals. CapFunc(1:0) in register RegCntConfig2 determines if the capture function is enabled or not and selects which edges of the capture signal source are valid for the capture operation. The source of the capture signal can be selected by setting CapSel(1:0) in the RegCntConfig2 register. For all sources, rising, falling or both edge sensitivity can be selected. Table 20-15 shows the capture condition as a function of the setting of these configuration bits. CapSel(1:0) Selected capture signal 11 1K 10 16 K 01 PA3 00 PA2 CapFunc 00 01 10 11 00 01 10 11 00 01 10 11 00 01 10 11 Selected condition Capture disabled Rising edge Falling edge Both edges Capture disabled Rising edge Falling edge Both edges Capture disabled Rising edge Falling edge Both edges Capture disabled Rising edge Falling edge Both edges Capture condition 1 K rising edge 1 K falling edge 2K 16 K rising edge 16 K falling edge 32 K PA3 rising edge PA3 falling edge PA3 both edges PA2 rising edge PA2 falling edge PA2 both edges Table 20-15: Capture condition selection CapFunc(1:0) and CapSel(1:0) can be modified only when the counters are stopped otherwise data may be corrupted during one counter clock cycle. Due to the synchronization mechanism of the shadow registers and depending on the frequency ratio between the capture and counter clocks, the interrupts may be generated one or only two counter clock pulses after the effective capture condition occurred. When the counters A and B are not cascaded and do not operate on the same clock, the interruptions on IrqA and IrqB which inform that the capture condition was met, may appear at different moments. In this case, the processor should read the shadow register associated to a counter only if the interruption related to this counter has been detected. An edge is detected on the capture signals only if the minimal pulse widths of these signals in the low and high states are higher than a period of the counter clock source. © Semtech 2006 www.semtech.com 20-10 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 20.12 Specifications Parameter Min Pulse width in the low and high states for an external clock source, frequency division by 2 disabled Pulse width in the low and high states for an external clock source, frequency division by 2 enabled Pulse width of external capture signals Typ Max Unit Conditions 500 ns @ 1.2V 125 ns @ 2.4V 100 ns @ 1.2V 25 ns @ 2.4V 1 fckcnt s Table 20-16: Timing specifications for the counters © Semtech 2006 www.semtech.com 20-11 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 21. The Voltage Level Detector 21.1 21.2 21.3 21.4 21.5 FEATURES ...................................................................................................................................... 21-2 OVERVIEW ...................................................................................................................................... 21-2 REGISTER MAP ................................................................................................................................ 21-2 INTERRUPT MAP .............................................................................................................................. 21-2 VLD OPERATION ............................................................................................................................. 21-3 © Semtech 2006 www.semtech.com 21-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 21.1 Features can be switched off, on or simultaneously with CPU activities generates an interrupt if power supply is below a pre-determined level 21.2 Overview The Voltage Level Detector monitors the state of the system battery. It returns a logical high value (an interrupt) in the status register if the supplied voltage drops below the user defined level (Vsb). 21.3 Register map There are two registers in the VLD, namely RegVldCtrl and RegVldStat. Table 21-2 shows the mapping of control bits and functionality of RegVldCtrl while Table 21-3 describes that for RegVldStat. register name RegVldCtrl RegVldStat Table 21-1: Vld registers pos. 7-4 3 RegVldCtrl -VldRange rw r rw reset 0000 0 nresetglobal 2-0 VldTune[2:0] rw 000 nresetglobal function reserved VLD detection voltage range for VldTune = “011”: 0 : 1.3V 1 : 2.55V VLD tuning: 000 : +19 % 111 : -18 % Table 21-2: RegVldCtrl pos. 7-3 2 RegVldStat -VldResult rw r r reset 00000 0 nresetglobal 1 0 VldValid VldEn r rw 0 nresetglobal 0 nresetglobal function reserved is 1 when battery voltage is below the detection voltage Indicates when VldResult can be read VLD enable Table 21-3: RegVldStat 21.4 Interrupt map interrupt source IrqVld default mapping in the interrupt manager RegIrqMid(2) Table 21-4: Interrupt map © Semtech 2006 www.semtech.com 21-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 21.5 VLD operation The VLD is controlled by VldRange, VldTune and VldEn. VldRange selects the voltage range to be detected, while VldTune is used to fine-tune this voltage level in 8 steps. VldEn is used to enable (disable) the VLD with a 1(0) value respectively. Disabled, the block will dissipate no power. symbol description min typ max unit Note 1 Vth TEOM TPW Threshold voltage VldRange 0 VldTune 1.53 1.44 0 001 1.36 0 010 1.29 0 011 1.22 0 100 1.16 0 101 1.11 0 110 0 111 3.06 1 000 2.88 1 001 2.72 1 010 2.57 1 011 2.44 1 100 2.33 1 101 2.22 1 110 2.13 1 111 V 1.06 duration of measurement Minimum pulse width detected comments trimming values: 2.0 2.5 ms Note 2 875 1350 us Note 2 000 Table 21-5: Voltage level detector operation Note 1: absolute precision of the threshold voltage is ±10%. Note 2: this timing is respected in case the internal RC or crystal oscillators are selected. Refer to the clock block documentation in case the external clock is used. To start the voltage level detection, the user sets bit VldEn. The measurement is started. After 2ms, the bit VldValid is set to indicate that the measurement results are valid. From that time on, as long as the VLD is enabled, a maskable interrupt request is sent if the voltage level falls below the threshold. One can also poll the VLD and monitor the actual measurement result by reading the VldResult bit of the RegVldStat. This result is only valid as long as the VldValid bit is ‘1’. © Semtech 2006 www.semtech.com 21-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver Figure 21-1 shows the timing of the VLD. An interrupt is generated on each rising edge of VldResult. vbat Vth vld_en vld_valid vld_result T EOM Τ PW Τ PW Figure 21-1: VLD timing The threshold value should not be changed during the measurement. © Semtech 2006 www.semtech.com 21-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 22. Low Power Comparators 22.1 22.2 22.3 22.4 FEATURES ............................................................................................................................22-2 OVERVIEW ............................................................................................................................22-2 REGISTER MAP ......................................................................................................................22-3 INTERRUPT MAP ....................................................................................................................22-4 © Semtech 2006 www.semtech.com 22-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 22.1 Features The cmpd peripheral implements four low power comparators. Quiescent current consumption of 1.5µA Very low switching current Per channel configurable interrupt Hysteresis 1 MHz operation 22.2 Overview Figure 22-1 gives an overview of this block: Cmpd 3 RegCmpdCtrl(7:5) 4 RegCmpdStat(7:4) IrqOnRisingCh (edge selection) 4 4 comparator output PB[7:4] comparators (analog) interrrupt enable EnIrqCh (channel enable) 5 RegCmpdCtrl(4:0) 4 RegCmpdStat(3:0) 4x Figure 22-1: Structure of Cmpd © Semtech 2006 www.semtech.com 22-2 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver The cmpd peripheral is a 4-channel low power comparator. It is intended to compare analog input signals with an internally set threshold voltage. The comparator maintains low current consumption even if the input signal is very close to the threshold. The comparison result of each channel can be used to generate an interrupt and/or is available for polling. The comparator can be enabled or disabled by programming the Enable bit in the RegCmpdCtrl register. When disabled, the block consumes no current. The peripheral has a single interrupt output which is a combination of the four channels. The combination can be chosen by programming the RegCmpdCtrl register. The EnIrqCh[3:0] bits select the channel that can activate the interrupt. The IrqOnRisingCh[2:0] bits indicate if the interrupt is generated on detection of the rising or falling edge of the channel. The comparison results of the peripheral can be read in the RegCmpdStat register. The bits CmpdOut[3:0] are the value of the comparisons at the moment the register is read. The CmpdStat[3:0] indicates which channel generated an interrupt since the register was last read. Comparator specifications: Sym tpulse IDDq IDDstat Vth ∆Vth/∆T Vhyst description min typ max Required input pulse width 500 Quiescent current 0.8 1.5 Maximal static current 1.5 Threshold voltage 0.7 1.1 Threshold temperature drift -0.9 Threshold hysteresis 13 Table 22-1: Comparator specifications unit ns µA µA V mV/°C mV comments VBAT≥1.2V 1 2 3 Comments: 1. The quiescent current is defined for a static input voltage <0.5V or >1.3V. The specified consumption is the sum for all 4 channels. 2. The maximal static current is defined for any static input voltage between VDD and VSS. The specified consumption is the sum for all 4 channels. 3. Defined with respect to VSS. How to start the cmpd: To avoid unwanted irqs one has first to configure the rising / falling edge of the detection (bit IrqOnRisingCh[2:0]) and to enable the comparator (bit Enable). Only after that may the user enable the channel interrupts with bit EnIrqCh[3:0]. 22.3 Register map There are two registers in the Cmpd, namely RegCmpdStat and RegCmpdCtrl. Table 22-3 and Table 22-4 show the mapping of the control bits and the functionality of these registers. register name RegCmpdStat RegCmpdCtrl Table 22-2: Cmpd registers © Semtech 2006 www.semtech.com 22-3 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver pos. 7 RegCmpdStat CmpdStat[3] rw rc reset 0 nresetglobal 6 CmpdStat[2] rc 0 nresetglobal 5 CmpdStat[1] rc 0 nresetglobal 4 CmpdStat[0] rc 0 nresetglobal 3 2 1 0 CmpdOut[3] CmpdOut[2] CmpdOut[1] CmpdOut[0] r r r r 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function 1: if the channel 3 generated an interrupt last read of this register 1: if the channel 2 generated an interrupt last read of this register 1: if the channel 1 generated an interrupt last read of this register 1: if the channel 0 generated an interrupt last read of this register Channel 3 comparator output Channel 2 comparator output Channel 1 comparator output Channel 0 comparator output since since since since Table 22-3: RegCmpdStat pos. 7 6 5 4 3 2 1 0 RegCmpdCtrl IrqOnRisingCh[2] IrqOnRisingCh[1] IrqOnRisingCh[0] EnIrqCh[3] EnIrqCh[2] EnIrqCh[1] EnIrqCh[0] Enable rw rw rw rw rw rw rw rw rw reset 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal 0 nresetglobal function 1: an interrupt is generated on the rising edge channels 2 and 3. 0: an interrupt is generated on the falling edge channels 2 and 3. 1: an interrupt is generated on the rising edge channel 1. 0: an interrupt is generated on the falling edge channel 1. 1: an interrupt is generated on the rising edge channel 0. 0: an interrupt is generated on the falling edge channel 0. 1 enables interrupt on channel 3 1 enables interrupt on channel 2 1 enables interrupt on channel 1 1 enables interrupt on channel 0 Enables the comparator of of of of of of Table 22-4: RegCmpdCtrl 22.4 Interrupt map interrupt source cmpd_irq default mapping in the interrupt manager RegIrqHigh[2] Table 22-5: Interrupt map © Semtech 2006 www.semtech.com 22-4 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver 23 Physical Dimensions 23.1 QFP type package The QFP package dimensions are given in Figure 23-1 and Table 23-1 Figure 23-1. QFP type package package LQFP-80 LQFP-100 A B C D E F mm 14.0 mm 14.0 mm 1.4 mm 0.10 mm 0.32 mm 0.65 14.0 14.0 1.4 0.10 0.22 0.5 Table 23-1. QFP package dimensions © Semtech 2006 www.semtech.com 23-1 XE8802 Sensing Machine Data Acquisition MCU with ZoomingADC™ and LCD driver © Semtech 2005 All rights reserved. Reproduction in whole or in part is prohibited without the prior written consent of the copyright owner. The information presented in this document does not form part of any quotation or contract, is believed to be accurate and reliable and may be changed without notice. No liability will be accepted by the publisher for any consequence of its use. Publication thereof does not convey nor imply any license under patent or other industrial or intellectual property rights. Semtech. assumes no responsibility or liability whatsoever for any failure or unexpected operation resulting from misuse, neglect improper installation, repair or improper handling or unusual physical or electrical stress including, but not limited to, exposure to parameters beyond the specified maximum ratings or operation outside the specified range. SEMTECH PRODUCTS ARE NOT DESIGNED, INTENDED, AUTHORIZED OR WARRANTED TO BE SUITABLE FOR USE IN LIFE-SUPPORT APPLICATIONS, DEVICES OR SYSTEMS OR OTHER CRITICAL APPLICATIONS. INCLUSION OF SEMTECH PRODUCTS IN SUCH APPLICATIONS IS UNDERSTOOD TO BE UNDERTAKEN SOLELY AT THE CUSTOMER’S OWN RISK. Should a customer purchase or use Semtech products for any such unauthorized application, the customer shall indemnify and hold Semtech and its officers, employees, subsidiaries, affiliates, and distributors harmless against all claims, costs damages and attorney fees which could arise. Contact Information Semtech Corporation Wireless and Sensing Products Division 200 Flynn Road, Camarillo, CA 93012 Phone (805) 498-2111 Fax : (805) 498-3804 © Semtech 2006 www.semtech.com 23-2