N CLC5902 Dual Digital Tuner/AGC 0 0 General Overview The CLC5902 Dual Digital Tuner/AGC IC is a two channel digital downconverter (DDC) with integrated automatic gain control (AGC). The CLC5902 is a key component in the Diversity Receiver Chipset (DRCS) which includes one CLC5902 Dual Digital Tuner/AGC, two CLC5956 12-bit analog-to-digital converters (ADCs), and two CLC5526 digitally controlled variable gain amplifiers (DVGAs). A block diagram for a Diversity Receiver Chipset based narrowband communications system is shown in Figure 1. This system allows direct IF sampling of signals up to 300MHz for enhanced receiver performance and reduced system costs. The CLC5902 offers high dynamic range digital tuning and filtering based on hard-wired digital signal processing (DSP) technology. Each channel has independent tuning, phase offset, and gain settings. Channel filtering is performed by a series of three filters. The first is a 4-stage Cascaded Integrator Comb (CIC) filter with a programmable decimation ratio from 8 to 2048. Next there are two symmetric FIR filters, a 21-tap and a 63-tap, both with programmable coefficients. The first FIR filter decimates the data by 2, the second FIR decimates by either 2 or 4. Channel filter bandwidth at 52MSPS ranges from ±650kHz down to ±1.3kHz. The CLC5902’s AGC controller monitors the ADC output and controls the ADC input signal level by adjusting the DVGA setting. AGC threshold, deadband+hysteresis, and the loop time constant are user defined. Total dynamic range of greater than 120dB fullscale signal to noise can be achieved with the Diversity Receiver Chipset. CLC5526 CLC5956 Features n 52MSPS Operation n Two Independent Channels with n n n n n n n n n n n n 14-bit inputs Greater than 100 dB image rejection Greater than 100 dB spurious free dynamic range 0.02 Hz tuning resolution User Programmable AGC Channel Filters include a Fourth Order CIC followed by 21-tap and 63-tap Symmetric FIRs FIR filters process 21-bit Data with 16-bit Programmable Coefficients Flexible output formats include 12-bit Floating Point or 8, 16, 24, and 32 bit Fixed Point Serial and Parallel output ports JTAG Boundary Scan 8-bit Microprocessor Interface 380mW/channel, 52 MHz, 3.3V 128 pin PQFP package Applications n Cellular Basestations n Satellite Receivers n Wireless Local Loop Receivers n Digital Communications CLC5902 LC IF A ADC DVGA 12 4 IF B DVGA ADC LC CLC5902 Dual Digital Tuner/AGC May 1999 12 Dual Digital Tuner/AGC SerialOutA/B SerialOutB SCK SFS RDY ParallelOutput[15..0] ParallelOutputEnable ParallelSelect[2..0] CLK Figure 1 Diversity Receiver Chipset Block Diagram ©1999 National Semiconductor Corporation Rev. 3.05 May 27, 1999 Channel A Controls AGC_IC_A GAIN_A AGC_RB_A FREQ_A PHASE_A DITH_A Microprocessor Interface RD WR CE A[7:0] D[7:0] AGAIN[2..0] ASTROBE AIN MUX A BIN MUX B 14 14 Channel A Tuning, Channel Filters, and AGC (see Figure 14) Channel B Tuning, Channel Filters, and AGC (see Figure 14) Output Formatter Floating Point: 4-bit Exponent and 8-bit Mantissa or Two’s Complement: 32-bit Truncated or 24-bit Rounded or 16-bit Rounded or 8-bit Truncated (see Figure 26) RDY POUT[15..0] PSEL[2..0] POUT_EN BSTROBE BGAIN[2..0] TEST_REG Channel B Controls AGC_IC_B GAIN_B AGC_RB_B FREQ_B PHASE_B DITH_B Input Source A_SOURCE B_SOURCE CK CLK GEN SI MR Sync Logic AGC_EN Figure 2 AOUT/BOUT BOUT SCK SFS Common Channel Controls DEC AGC_FORCE DEC_BY_4 AGC_RESET_EN SCALE AGC_HOLD_IC EXP_INH AGC_LOOP_GAIN F1_COEFF AGC_COUNT F2_COEFF AGC_TABLE Output Controls RATE SOUT_EN SCK_POL SFS_POL RDY_POL MUX_MODE PACKED FORMAT DEBUG_EN DEBUG_TAP CLC5902 Dual Digital Tuner/AGC Block Diagram with Control Register Associations Functional Description. The CLC5902 block diagram is shown in Figure 2. The CLC5902 contains two identical digital down-conversion (DDC) circuits. Each DDC accepts a 14-bit sample at up to 52MSPS, down converts from a selected carrier frequency to baseband, decimates the signal rate by a programmable factor ranging from 32 to 16384, provides channel filtering, and outputs quadrature symbols. A crossbar switch enables either of the two inputs or a test register to be routed to either DDC channel. Flexible channel filtering is provided by the two programmable decimating FIR filters. The final filter outputs can be converted to a 12-bit floating point format or standard two’s complement format. The output data is available at both serial and parallel ports. The CLC5902 maintains over 100 dB of spurious free dynamic range and over 100 dB of out-of-band rejection. This allows considerable latitude in channel filter partitioning between the analog and digital domains. The frequencies, phase offsets, and phase dither of the two sine/cosine numerically controlled oscillators (NCOs) can be independently specified. Both channels share the same Rev. 3.05 May 27, 1999 2 decimation ratio, bandwidth, filter coefficients, and input/ output formats. Each channel has its own AGC circuit for use with narrowband radio channels where most of the channel filtering precedes the ADC. The AGC closes the loop around the CLC5526 DVGA, compressing the dynamic range of the signal into the ADC. The AGC can be configured to operate continuously or in a gated mode. The two AGC circuits operate independently but share the same programmed parameters and control signals. The chip receives configuration and control information over a microprocessor-compatible bus consisting of an 8bit data I/O port, an 8-bit address port, a chip enable strobe, a read strobe, and a write strobe. The chip’s control registers (8 bits each) are memory mapped into the 8-bit address space of the control port. JTAG boundary scan and on-chip diagnostic circuits are provided to simplify system debug and test. The CLC5902 supports 3.3V I/O. The CLC5956 ADC outputs are compatible with the CLC5902 inputs. The CLC5902 outputs swing to the 3.3V rail so they can be directly connected to 5V TTL inputs if desired. ©1999 National Semiconductor Corporation CLC5902 Electrical Characteristics (VCC=+3.3V, 52MHz, CIC Decimation=48, F2 Decimation=2, Tmin=-40°C, Tmax=+85°C; unless specified) DC Characteristics PARAMETER SYMBOL MIN TYP MAX UNITS Notes Voltage input low VIL -0.5 0.8 V 1 Voltage input high VIH 2.0 VCC+0.5 V 1 Input current IIN 10 uA 1 Voltage output low (IOL = 4mA/12mA, see pin description) VOL 0.4 V 1 Voltage output high (IOH = -4mA/-12mA, see pin description) VOH V 1 Input capacitance CIN 4.0 pF 3 2.4 AC Characteristics PARAMETER (CL=50pF) SYMBOL MIN TYP MAX UNITS Clock (CK) Frequency (Figure 7) FCK Spurious Free Dynamic Range SFDR -100 dBFS Signal to Noise Ratio SNR -127 dBFS Tuning Resolution 0.02 Hz Phase Resolution 0.005 ° 52 MHz Notes 1 MR Active Time (Figure 5) tMRA 4 CK periods 1 MR Inactive to first Control Port Access (Figure 5) tMRIC 10 CK periods 1 MR Setup Time to CK (Figure 5) tMRSU 9 ns 1 MR Hold Time to CK (Figure 5) tMRH 2 ns 1 MR Inactive to A|BSTROBE Release (Figure 5) tMRSR SI Setup Time to CK (Figure 6) tSISU 9 ns 1 SI Hold Time from CK (Figure 6) tSIH 2 ns 1 SI Pulse Width (Figure 6) tSIW 4 CK periods 1 SI Inactive to A|BSTROBE Release (Figure 6) tSISR CK duty cycle (Figure 7) tCKDC 17 ns 17 40 ns 60 % 1 CK rise and fall times (VIL to VIH) (Figure 7) tRF ns 1 Input setup before CK goes high (A|BIN) (Figure 7) tSU 7 ns 1 Input hold time after CK goes high (Figure 7) tHD 3 ns 1 A|BSTROBE Pulse Width (Figure 8) tSTBPW 1 CK period 2 1 CK period 2 CK periods 1 3 A|BGAIN Valid Setup before A|BSTROBE (Figure 8) tGSU AGC_EN Active Width (Figure 8) tENW 2 SCK to SFS Valid (Table Note A) (Figure 9) tSFSV 0 7 ns 1 SCK to A|BOUT Valid (Table Note B) (Figure 9) tOV 0 7 ns 1 RDY Pulse Width (Figure 9) tRDYW CK periods 1 POUT_EN Active to POUT[15..0] Valid (Figure 10) tOENV 15 ns 1 POUT_EN Inactive to POUT[15..0] Tri-State (Figure 10) tOENT 15 ns 1 PSEL[2..0] to POUT[15..0] Valid (Figure 11) tSELV 20 ns 1 RDY to POUT[15..0] New Value Valid (Table Note C) (Figure 12) tRDYV 10 ns 1 Propagation Delay TCK to TDO (Figure 13) tPLH 30 ns 1 Propagation Delay TCK to Data Out (Figure 13) tPHL 35 ns 1 Disable Time TCK to TDO (Figure 13) tPLZ 35 ns 1 Disable Time TCK to Data Out (Figure 13) tPHZ Enable Time TCK to TDO (Figure 13) tPZL ©1999 National Semiconductor Corporation 3 4 0 35 ns 1 35 ns 1 Rev. 3.05 May 27, 1999 PARAMETER (CL=50pF) SYMBOL MIN Enable Time TCK to Data Out (Figure 13) tPZH 0 Setup Time Data to TCK (Figure 13) tS Setup Time TDI to TCK (Figure 13) TYP MAX 35 UNITS Notes ns 1 10 ns 1 tS 10 ns 1 Setup Time TMS to TCK (Figure 13) tS 15 ns 1 Hold Time Data to TCK (Figure 13) tH 55 ns 1 Hold Time TCK to TDI (Figure 13) tH 55 ns 1 Hold Time TCK to TMS (Figure 13) tH 10 ns 1 TCK Pulse Width High (Figure 13) tWH 55 ns 1 TCK Pulse Width Low (Figure 13) tWL 40 ns 1 TCK Maximum Frequency (Figure 13) JTAGFMAX MHz 1 Control Setup before the controlling signal goes low (Figure 14) tCSU 5 ns 1 Control hold after the controlling signal goes high (Figure 14) tCHD 5 ns 1 Controlling strobe pulse width (Write) (Figure 14) tCSPW 30 ns 1 Control output delay controlling signal low to D (Read) (Figure 14) tCDLY 30 ns 1 Control tri-state delay after controlling signal goes high (Figure 14) tCZ 20 ns 1 Dynamic Supply Current (FCK =52MHz, N=48) ICC 230 280 mA 1 Dynamic Supply Current (FCK =52MHz, N=8) ICC 260 320 mA 1 10 essarily implied. Exposure to maximum ratings for extended periods may affect device reliability. Notes A - C A. tSFSV refers to the rising edge of SCK when SCK_POL=0 and the falling edge when SCK_POL=1. B. tOV refers to the rising edge of SCK when SCK_POL=0 and the falling edge when SCK_POL=1. C. tRDYV refers to the rising edge of RDY when RDY_POL=0 and the falling edge when RDY_POL=1. Notes 1 - 3 Recommended Operating Conditions Positive Supply Voltage (VCC) 3.3V ±10% Operating Temperature Range -40°C to +85°C Package Thermal Resistance 1. These parameters are 100% tested at 25°C. 2. Typical specifications are the mean values of the distributions of deliverable CLC5902s tested to date. 3. Min/max ratings are based on product characterization and simulation. Individual parameters are tested as noted. Outgoing quality levels are determined from tested parameters. θja Package 128 pin PQFP 39°C/W θjc TBD°C/W Reliability Information Absolute Maximum Ratings Positive Supply Voltage (VCC) Voltage on Any Input or Output Pin Transistor Count -0.3V to 4.2V -0.3V to VCC+0.5V Input Current at Any Pin ±25mA Package Input Current ±50mA Package Dissipation at TA=25°C Ordering Information Order Code 1W ESD Susceptibility Human Body Model Machine Model 1500V 100V Soldering Temperature, Infrared, 10 seconds 300°C Storage Temperature CLC5902VLA 4 Temperature Range -40°C to +85°C Description 128-pin PQFP (industrial temperature range) CLC-DRCS-PCASM Fully loaded Diversity Receiver Chipset evaluation board and control panel software. CLC-CAPT-PCASM Data Capture board for the DRCS with Matlab analysis routines. -65°C to 150°C NOTE: Absolute maximum ratings are limiting values, to be applied individually, and beyond which the serviceability of the circuit may be impaired. Functional operability under any of these conditions is not nec- Rev. 3.05 May 27, 1999 1.2 million ©1999 National Semiconductor Corporation Package Dimensions DETAIL A Dimension are in millimeters Figure 3 CLC5902 Package Dimensions ©1999 National Semiconductor Corporation 5 Rev. 3.05 May 27, 1999 CLC5902 Dual Digital Tuner/AGC (Top View) 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 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 38 102 101 100 99 98 97 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 NC NC NC NC VSS POUT[5] POUT[6] POUT[7] POUT[8] POUT[9] VDD POUT[10] POUT[11] VSS POUT[12] POUT[13] POUT[14] VDD POUT[15] VSS AOUT SFS SCK VDD BOUT RDY VSS D[0] VDD D[1] D[2] D[3] D[4] D[5] VSS NC NC NC VDD BGAIN[2] BGAIN[1] BGAIN[0] BSTROBE AGC_EN MR SI VSS A[7] VDD A[6] VSS A[5] A[4] A[3] A[2] A[1] A[0] WR RD CE VSS D[7] D[6] VDD NC NC VSS (MSB) AIN[13] AIN[12] AIN[11] AIN[10] AIN[9] AIN[8] AIN[7] VDD AIN[6] AIN[5] AIN[4] AIN[3] AIN[2] AIN[1] AIN[0] VSS CK VDD (MSB) BIN[13] BIN[12] BIN[11] BIN[10] BIN[9] BIN[8] BIN[7] VSS BIN[6] BIN[5] BIN[4] BIN[3] BIN[2] BIN[1] BIN[0] VSS NC 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 VDD AGAIN[2] AGAIN[1] AGAIN[0] ASTROBE NC SCAN_EN TRST VSS TCK TMS TDI TDO VDD POUT_SEL[0] POUT_SEL[1] POUT_SEL[2] POUT_EN VSS POUT[0] POUT[1] VDD POUT[2] POUT[3] POUT[4] VSS CLC5902 Pinout Figure 4 CLC5902 Pinout Rev. 3.05 May 27, 1999 6 ©1999 National Semiconductor Corporation Pin Descriptions Signal Pin DESCRIPTION MR 45 MASTER RESET, Active low Resets all registers within the chip. ASTROBE and BSTROBE are asserted during MR. AIN[13:0], BIN[13:0] 4:10,12:18 22:28,30:36 INPUT DATA, Active high 2’s complement input data. AIN[13] and BIN[13] are the MSBs. The data is clocked into the chip on the rising edge of the clock (CK). The CLC595X connects directly to these input pins with no additional logic. SERIAL OUTPUT DATA, Active high The 2’s complement serial output data is transmitted on these pins, MSB first. The output bits change on the rising edge of SCK (falling edge if SCK_POL=1) and should be captured on the falling edge of SCK (rising if SCK_POL=1). These pins are tri-stated at power up and are enabled by the SOUT_EN control register bit. See Figure 9 and Figure 29 timing diagrams. In Debug Mode AOUT=DEBUG[1], BOUT=DEBUG[0]. AOUT BOUT (12mA drive) 82 78 AGAIN[2:0], BGAIN[2:0] 125:127 40:42 OUTPUT DATA TO DVGA, Active high 3 bit bus that sets the gain of the DVGA determined by the AGC circuit. ASTROBE, BSTROBE 124 43 DVGA STROBE, Active low Strobes the data into the DVGA. See Figure 8 and Figure 33 timing diagrams. SCK (12mA drive) 80 SFS (12mA drive) 81 SERIAL DATA CLOCK, Active high or low The serial data is clocked out of the chip by this clock. The active edge of the clock is user programmable. This pin is tri-stated at power up and is enabled by the SOUT_EN control register bit. See Figure 9 and Figure 29 timing diagrams. In Debug Mode outputs an appropriate clock for the debug data. SERIAL FRAME STROBE, Active high or low The serial word strobe. This strobe delineates the words within the serial output streams. This strobe is a pulse at the beginning of each serial word (PACKED=0) or each serial word I/Q pair (PACKED=1). The polarity of this signal is user programmable. This pin is tri-stated at power up and is enabled by the SOUT_EN control register bit. See Figure 9 and Figure 29 timing diagrams. In Debug Mode SFS=DEBUG[2]. POUT[15:0] (12mA drive) 84,86:88,90,91, 93:97,104:106, 108,109 PARALLEL OUTPUT DATA, Active high The output data is transmitted on these pins in parallel format. The POUT_SEL[2..0] pins select one of eight 16-bit output words. The POUT_EN pin enables these outputs. POUT[15] is the MSB. In Debug Mode POUT[15..0]=DEBUG[19..4]. POUT_SEL[2:0] 112:114 PARALLEL OUTPUT DATA SELECT, Active high The 16-bit output word is selected with these 3 pins according to Table 3. Not used in Debug Mode. POUT_EN 111 PARALLEL OUTPUT ENABLE. Active low This pin enables the chip to output the selected output word on the POUT[15:0] pins. Not used in Debug Mode. READY FLAG, Active high or low The chip asserts this signal to identify the beginning of an output sample period (OSP). The polarity of this signal is user programmable. This signal is typically used as an interrupt to a DSP chip, but can also be used as a start pulse to dedicated circuitry. This pin is active regardless of the state of SOUT_EN. In Debug Mode RDY=DEBUG[3]. RDY (12mA drive) 77 CK 20 INPUT CLOCK. Active high The clock input to the chip. The AIN, BIN, and SI input signals are clocked into the chip on the rising edge of this clock. SI 46 SYNC IN. Active low The sync input to the chip. The decimation counters, dither, and NCO phase can be synchronized by SI. This sync is clocked into the chip on the rising edge of the input clock (CK). Tie this pin high if external sync is not required. All sample data is flushed by SI. To properly initialize the DVGA ASTROBE and BSTROBE are asserted during SI. D[7:0] (12mA drive) 62,63,69:73,75 Table 1 DATA BUS. Active high This is the 8 bit control data I/O bus. Control register data is loaded into the chip or read from the chip through these pins. The chip will only drive output data on these pins when CE is low, RD is low, and WR is high. CLC5902 Pin Descriptions ©1999 National Semiconductor Corporation 7 Rev. 3.05 May 27, 1999 Signal Pin DESCRIPTION A[7:0] 48,50,52:57 ADDRESS BUS. Active high These pins are used to address the control registers within the chip. Each of the control registers within the chip are assigned a unique address in a flat address space. A control register can be written to or read from by setting A[7:0] to the register’s address. RD 59 READ ENABLE. Active low This pin enables the chip to output the contents of the selected register on the D[7:0] pins when CE is also low. WR 58 WRITE ENABLE. Active low This pin enables the chip to write the value on the D[7:0] pins into the selected register when CE is also low. This pin can also function as RD/WR if RD is held low. CE 60 CHIP ENABLE. Active low This control strobe enables the read or write operation. The contents of the register selected by A[7:0] will be output on D[7:0] when RD is low and CE is low. If WR is low and CE is low, then the selected register will be loaded with the contents of D[7:0]. AGC_EN 44 AGC ENABLE. Active low When enabled this pin starts the AGC counter. The AGC will operate until the counter decrements to zero then the AGC holds the last setting. TDO 116 TEST DATA OUT. Active high TDI 117 TEST DATA IN. Active high with pull-up TMS 118 TEST MODE SELECT. Active high with pull-up TCK 119 TEST CLOCK. Active high TRST 121 RESET. Active low with pull-up Asynchronous reset for TAP controller. Tie low or to MR if JTAG is not used. SCAN_EN 122 SCAN ENABLE. Active low with pull-up Enables access to internal scan registers. Tie high. Used for manufacturing test only! VSS 3,19,29,37,47,51, 61,68,76,83,89, 98,103,110,120 Ground. Quantity 15. VDD 11,21,39,49,64, 74,79,85,92,107, 115,128 Power. Quantity 12. Table 1 CLC5902 Pin Descriptions Timing Diagrams CK tMRH tMRSU tMRA MR tMRIC RD or WR tMRSR A|BSTROBE Figure 5 CLC5902 Master Reset Timing Rev. 3.05 May 27, 1999 8 ©1999 National Semiconductor Corporation CK tSIH tSISU tSIW SI tSISR A|BSTROBE Figure 6 CLC5902 Synchronization Input (SI) Timing 1/FCK tCKDC tCKDC CK VIH VIL tHD tRF tSU A|BIN Figure 7 CLC5902 ADC Input and Clock Timing CK tSTBPW A|BSTROBE tGSU (n-1) A|BGAIN[2..0] valid (n) tENW AGC_EN Figure 8 CLC5902 DVGA Interface Timing ©1999 National Semiconductor Corporation 9 Rev. 3.05 May 27, 1999 SCK SCK_POL=0 SCK=CK/2 SCK SCK_POL=1 tSFSV tSFSV SFS SFS_POL=0 SFS SFS_POL=1 tOV tOV lsb or undef A|BOUT msb Previous Q Output Word msb-1 msb-2 msb-3 I Output Word RDY RDY_POL=0 tRDYW RDY RDY_POL=1 Figure 9 CLC5902 Serial Port Timing POUT_EN tOENV tOENT POUT[15..0] Figure 10 CLC5902 Parallel Output Enable Timing POUT_SEL[2..0] n n+1 n+2 tSELV POUT[15..0] Figure 11 tSELV output (n) output (n+1) output (n+2) CLC5902 Parallel Output Select Timing RDY RDY_POL=0 RDY RDY_POL=1 tRDYV POUT[15..0] Figure 12 old output new output CLC5902 Parallel Output Data Ready Timing Rev. 3.05 May 27, 1999 10 ©1999 National Semiconductor Corporation TCK tPLH,tPHL TDO, D TCK tPZH tPHZ tPZL tPLZ D TCK TDO tWH tWL 1/JTAGfmax TCK tH tS TMS, TDI, D Figure 13 CLC5902 JTAG Port Timing CE WR tCHD RD tCSU A[7:0] tCDLY tCZ D[7:0] READ CYCLE; NORMAL MODE CE tCSPW WR RD tCSU A[7:0] tCHD D[7:0] WRITE CYCLE; NORMAL MODE CE tCHD WR tCSU A[7:0] tCDLY tCZ D[7:0] READ CYCLE; RD HELD LOW tCSPW CE WR tCSU A[7:0] tCHD D[7:0] WRITE CYCLE; RD HELD LOW Figure 14 CLC5902 Control I/O Timing ©1999 National Semiconductor Corporation 11 Rev. 3.05 May 27, 1999 tialized, the NCO dither generators will be reset, and the CIC decimation ratio will be initialized. Only the configuration data loaded into the microprocessor interface remains unaffected. Detailed Description Control Interface The CLC5902 is configured by writing control information into 148 control registers within the chip. The contents of these control registers and how to use them are described in Table 5. The registers are written to or read from using the D[7:0], A[7:0], CE, RD and WR pins (see Table 1 for pin descriptions). This interface is designed to allow the CLC5902 to appear to an external processor as a memory mapped peripheral. See Figure 14 for details. SI may be held low as long as desired after a minimum of 4 CK periods. Input Source The input crossbar switch allows either AIN, BIN, or a test register to be routed to the channel A or channel B AGC/DDC. The AGC outputs, AGAIN and BGAIN, are not switched. If AIN and BIN are exchanged the AGC loop will be open and the AGCs will not function properly. AIN and BIN should meet the timing requirements shown in Figure 7. The control interface is asynchronous with respect to the system clock, CK. This allows the registers to be written or read at any time. In some cases this might cause an invalid operation since the interface is not internally synchronized. In order to assure correct operation, SI must be asserted after the control registers are written. Selecting the test register as the input source allows the AGC or DDC operation to be verified with a known input. See the test and diagnostics section for further discussion. The D[7:0], A[7:0], WR, RD and CE pins should not be driven above the positive supply voltage. Down Converters Master Reset A detailed block diagram of each DDC channel is shown in Figure 15. Each down converter uses a complex NCO and mixer to quadrature downconvert a signal to baseband. The “FLOAT TO FIXED CONVERTER” treats the 15-bit mixer output as a mantissa and the AGC output, EXP, as a 3-bit exponent. It performs a bit shift on the data based on the value of EXP. This bit shifting is used to expand the compressed dynamic range resulting from the DVGA operation. The DVGA gain is adjusted in 6dB steps which are equivalent to each digital bit shift. A master reset pin, MR, is provided to initialize the CLC5902 to a known condition and should be strobed after power up. This signal will clear all sample data and all user programmed data (filter coefficients and AGC settings). All outputs will be disabled (tri-stated). ASTROBE and BSTROBE will be asserted to initialize the DVGA values. Table 5 describes the control register default values. Synchronizing Multiple CLC5902 Chips The exponent (EXP) can be forced to its maximum value by setting the EXP_INH bit. If x in ( n ) is the DDC input, the signal after the “FLOAT TO FIXED CONVERTER” is A system containing two or more CLC5902 chips will need to be synchronized if coherent operation is desired. To synchronize multiple CLC5902 chips, connect all of the sync input pins together so they can be driven by a common sync strobe. Synchronization occurs on the rising edge of CK when SI goes back high. When SI is asserted all sample data will be flushed immediately, the numerically controlled oscillator (NCO) phase offset will be ini- 17 SIN COS x3 ( n ) Data @ FCK/N DEC_BY_4 F2_COEF Data @ FCK/N*2 21 I SAT 21 F2 FILTER DECIMATE BY 2 OR 4 21 SAT & ROUND 21 F1 FILTER DECIMATE BY 2 22 SHIFT UP 22 SAT & ROUND 15 17 F1_COEF DEC CIC FILTER DECIMATE BY 8 TO 2K Data @ FCK = FS (FSAMPLE) GAIN_A SCALE SHIFT UP 14 15 FLOAT TO FIXED CONVERTER ROUND EXPONENT MUXA EQ. 1 EXP 3 x in ( n ) EXP for the I component. Changing the ‘cos’ to ‘sin’ in this equation will provide the Q component. EXP_INH EXP (from AGC) FREQ_A x 3 ( n ) = x in ( n ) • cos ( ωn ) • 2 TO OUTPUT CIRCUIT Q Data @ FCK/N*2*F2_DEC = OFS (Output FSAMPLE) NCO PHASE_A Figure 15 N = DEC + 1 CLC5902 Down Converter, Channel A (Channel B is identical) Rev. 3.05 May 27, 1999 12 ©1999 National Semiconductor Corporation The “FLOAT TO FIXED CONVERTER” circuit expands the dynamic range compression performed by the DVGA. Signals from this point onward extend across the full dynamic range of the signals applied to the DVGA input. This allows the AGC to operate continuously through a burst without producing artifacts in the signal due to the settling response of the decimation filters after a 6dB DVGA gain adjustment. For example, if the DVGA input signal were to increase causing the ADC output level to cross the AGC threshold level, the gain of the DVGA would change by -6dB. The 6dB step is allowed to propagate through the ADC and mixers and is compensated out just before the filtering. The accuracy of the compensation is dependent on the accuracy of the DVGA gain step. This operating mode requires 21 bits (14-bit ADC output + 7bit shift) to represent the full linear dynamic range of the signal. The output word must be set to either 24-bit or 32bit to take advantage of the entire dynamic range available. The CLC5902 can also be configured to output a floating point format with up to 138dB of numerical resolution using only 12 output bits. The “SHIFT UP” circuit will be discussed in the Four Stage CIC filter section on page 14. A 4-stage cascaded-integrator-comb (CIC) filter and a two-stage decimate by 4 or 8 finite impulse response (FIR) filter are used to lowpass filter and isolate the desired signal. The CIC filter reduces the sample rate by a programmable factor ranging from 8 to 2048 (decimation ratio). The CIC outputs are followed by a gain stage and then followed by a two-stage decimate by 4 or 8 filter. The gain circuit allows the user to boost the gain of weak signals by up to 42 dB in 6 dB steps. It also rounds the signal to 21 bits and saturates at plus or minus full scale. The first stage of the two stage filter is a 21-tap, symmetric decimate by 2 FIR filter (F1) with programmable 16 bit tap weights. The coefficients of the first 11 taps are downloaded to the chip as 16 bit words. Since the filter is a symmetric configuration only the first 11 coefficients must be loaded. The F1 section on page 15 provides a generic set of coefficients that compensate for the rolloff of the CIC filter and provide a passband flat to 0.01dB with 70 dB of out of band rejection. A second coefficient set is provided that has a narrower output passband and greater out-ofband rejection. The second set of coefficients is ideal for systems such as GSM where far-image rejection is more important than adjacent channel rejection. The second stage is a 63 tap decimate by 2 or 4 programmable FIR filter (F2) also with 16 bit tap weights. Filter coefficients for a flat response from -0.4FS to +0.4FS of the output sample rate with 80dB of out of band rejection are provided in the F2 section. A second set of F2 coefficients is also provided to enhance performance for GSM systems. The user can also design and download their own final filter to customize the channel’s spectral response. Typical uses of programmable filter F2 include matched (root-raised cosine) filtering, or filtering to generate oversampled outputs with greater out of band rejection. The 63 tap symmetrical filter is downloaded into the chip as 32 words, 16 bits each. Saturation to plus or minus full scale is performed at the output of F1 and F2 to clip the signal rather than allow it to roll over. The Numerically Controlled Oscillator The tuning frequency of each down converter is specified as a 32 bit word (.02Hz resolution at CK=52MHz) and the phase offset is specified as a 16 bit word (.005°). These two parameters are applied to the Numerically Controlled Complex NCO Output 0 0 -20 -20 -40 -40 Magnitude (dB) Magnitude (dB) Complex NCO Output -60 -80 -60 -80 -100 -100 -120 -120 -140 -140 -160 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 -160 -0.5 0.5 Frequency Normalized to FS -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 Frequency Normalized to FS (b) After Phase Dithering (a) Before Phase Dithering Figure 16 -0.4 Example of NCO spurs due to phase truncation ©1999 National Semiconductor Corporation 13 Rev. 3.05 May 27, 1999 Oscillator (NCO) circuit to generate sine and cosine signals used by the digital mixer. The NCOs can be synchronized with NCOs on other chips via the sync pin SI. This allows multiple down converter outputs to be coherently combined, each with a unique phase and amplitude. Complex NCO Output Phase Dither Disabled 0 NCO frequency swept through FS/8 from -FCK/2 to +FCK(1-2-31)/2. In some cases the sampling process causes the order of the I and Q components to be reversed. Should this occur simply invert the polarity of the tuning frequency F. -20 -40 Magnitude (dB) The tuning frequency is set by loading the FREQ register according to the formula FREQ = 232F/FCK, where F is the desired tuning frequency and FCK is the chip’s clock rate. FREQ is a 2’s complement word. The range for F is -60 -80 -100 -120 -140 The 2’s complement format represents full-scale negative as 10000000 and full-scale positive as 01111111 for an 8bit example. is the desired phase in radians ranging between 0 and 2π. PHASE is an unsigned 16-bit number. P ranges from 0 to 2π(1-2-16). Phase dithering can be enabled to reduce the spurious signals created by the NCO due to phase truncation. This truncation is unavoidable since the frequency resolution is much finer than the phase resolution. With dither enabled, spurs due to phase truncation are below -100 dBc for all frequencies and phase offsets. Each NCO has its own dither source and the initial state of one is maximally offset with respect to the other so that they are effectively uncorrelated. The phase dither sources are on by default. They are independently controlled by the DITH_A and DITH_B bits. The amplitude resolution of the ROM creates a worst-case spur amplitude of -101dBc rendering amplitude dither unnecessary. The spectrum plots in Figure 16 show the effectiveness of phase dither in reducing NCO spurs due to phase truncation for a worst-case example (just below F S /8). With dither off, the spur is at -86.4dBFS. With dither on, the spur is below -125dBFS, disappearing into the noise floor. This spur is spread into the noise floor which results in an SNR of -83.6dBFS. Figure 17 shows the spur levels as the tuning frequency is scanned over a narrow portion of the frequency range. The spurs are again a result of phase quantization but their locations move about as the frequency scan progresses. As before, the peak spur level drops when dithering is enabled. When dither is enabled and the fundamental frequency is exactly at F S /8, the worst-case spur due to amplitude quantization can be observed at -101dBc in Figure 18. Rev. 3.05 May 27, 1999 14 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 Frequency Normalized to FS Figure 17 NCO Spurs due to Phase Quantization Complex NCO Output Phase Dither Enabled 0 -20 -40 Magnitude (dB) The 16 bit phase offset is set by loading the PHASE register according to the formula PHASE = 216P/2π, where P -160 -0.5 -60 -80 -100 -120 -140 -160 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 Frequency Normalized to FS Figure 18 Worst Case Amplitude Spur, NCO at FS/8 Four Stage CIC Filter The mixer outputs are decimated by a factor of N in a four stage CIC filter. N is programmable to any integer between 8 and 2048. Decimation is programmed in the DEC register where DEC = N - 1. The programmable decimation allows the chip’s usable output bandwidth to range from about 2.6kHz to 1.3MHz when the input data rate (which is equal to the chip’s clock rate, FCK) is 52 MHz. A block diagram of the CIC filter is shown in Figure 19. ©1999 National Semiconductor Corporation Data @ FCK = FS (FSAMPLE) Figure 19 DECIMATE BY FACTOR OF N 22 66 22-bit input to SHIFT_UP is aligned at the bottom of the 66-bit path when SCALE=0. DATA OUT Data @ FCK/N N = DEC + 1 Four-stage decimate by N CIC filter The CIC filter has a gain equal to N4 (filter decimation^4) which must be compensated for in the “SHIFT UP” circuit shown in Figure 19 as well as Figure 15. This circuit has a gain equal to 2(SCALE-44), where SCALE ranges from 0 to 40. This circuit divides the input signal by 244 providing maximum headroom through the CIC filter. For optimal noise performance the SCALE value is set to increase this level until the CIC filter is just below the point of distortion. A value is normally calculated and loaded for SCALE such that GAIN SHIFTUP ⋅ GAIN CIC ≤ 1 . The actual gain of the CIC filter will only be unity for powerof-two decimation values. In other cases the gain will be somewhat less than unity. Channel Gain symmetric with the following 11 unique taps (1|21, 2|20, ..., 10|12, 11): 29, -85, -308, -56, 1068, 1405, -2056, -6009, 1303, 21121, 32703 Frequency Response of F1 Using STD Set 0 −10 −20 −30 Magnitude (dB) DATA IN SHIFT UP SCALE −40 −50 −60 −70 −80 The gain of each channel can be boosted up to 42 dB by shifting the output of the CIC filter up by 0 to 7 bits prior to rounding it to 21 bits. For channel A, the gain of this −90 −100 0 0.1 0.2 0.3 0.4 Frequency Normalized To Filter Input Sample Rate GAIN_A stage is: GAIN = 2 , where GAIN_A ranges from 0 to 7. Overflow due to the GAIN circuit is saturated (clipped) at plus or minus full scale. Each channel can be given its own GAIN setting. First Programmable FIR Filter (F1) The CIC/GAIN outputs are followed by two stages of filtering. The first stage is a 21 tap decimate-by-2 symmetric FIR filter with programmable coefficients. Typically, this filter compensates for a slight droop induced by the CIC filter. In addition, it often provides stopband assistance to F2 when deep stop bands are required. The filter coefficients are 16-bit 2’s complement numbers. Unity gain will be achieved through the filter if the sum of the 21 coefficients is equal to 216. If the sum is not 216, then F1 will introduce a gain equal to (sum of coefficients)/216. The 21 coe ffi ci ent s a re ide nti fi ed as c oeffic ie nts h 1 ( n ), n = 0, …, 20 where h 1 ( 10 ) is the center tap. The coefficients are symmetric, so only the first 11 are loaded into the chip. Figure 20 0.5 F1 STD frequency response The second set of coefficients (GSM set) are intended for applications that need deeper stop bands or need oversampled outputs. These requirements are common in cellular systems where out of band rejection requirements can exceed 100dB (see Figure 21). They are useful for wideband radio architectures where the channelization is done after the ADC. These filter coefficients introduce a gain of 0.984 and are: -49, -340, -1008, -1617, -1269, 425, 3027, 6030, 9115, 11620, 12606 Second Programmable FIR Filter (F2) The second stage decimate by two or four filter also uses externally downloaded filter coefficients. The filter coefficients are 16-bit 2’s complement numbers. Unity gain will be achieved through the filter if the sum of the 63 coefficients is equal to 216. If the sum is not 216, then the F2 will introduce a gain equal to (sum of coefficients)/216. Two example sets of coefficients are provided here. The first set of coefficients, referred to as the standard set (STD), compensates for the droop of the CIC filter providing a passband which is flat (0.01 dB ripple) over 95% of the final output bandwidth with 70dB of out-of-band rejection (see Figure 20). The filter has a gain of 0.999 and is The 63 coefficients are identified as h 2 ( n ) , n = 0 ,… ,62 15 Rev. 3.05 May 27, 1999 ©1999 National Semiconductor Corporation where h 2 ( 31 ) is the center tap. The coefficients are symmetric, so only the first 32 are loaded into the chip. Frequency Response of F1 Using GSM Set Frequency Response of F2 Using GSM Set 5 0 0 −5 −20 Magnitude (dB) Magnitude (dB) −10 −40 −60 −15 −20 −25 −30 −80 −35 −40 −100 −45 −120 0 0.1 0.2 0.3 0.4 Frequency Normalized To Filter Input Sample Rate Figure 21 −50 0.5 F1 GSM frequency response 0 Figure 23 An example filter (STD F2 coefficients, see Figure 22) with 80dB out-of-band rejection, gain of 1.00, and 0.03 dB peak to peak passband ripple is created by this set of 32 unique coefficients: -14, -20, 19, 73, 43, -70, -82, 84, 171, -49, -269, 34, 374, 192, -449, 0.1 0.2 0.3 0.4 Frequency Normalized To Filter Input Sample Rate 0.5 F2 GSM frequency response The filter coefficients of both filters can be used to tailor the spectral response to the user’s needs. For example, the first can be loaded with the standard set to provide a flat response through to the second filter. The latter can then be programmed as a Nyquist (typically a root-raisedcosine) filter for matched filtering of digital data. Combined Frequency Response of CIC/F1/F2 Using STD Set -430, 460,751, -357, -1144, 81, 1581, 443, -2026, -1337, 2437, 2886, 0 FCK = 52MHz Decimation = 192 OFS = 270.83kHz -2770, -6127, 2987, 20544, 29647 Magnitude (dB) Frequency Response of F2 Using STD Set 0 −10 −20 Magnitude (dB) −30 −50 −100 −40 −50 −60 −150 −70 0 −80 500 1000 1500 2000 Frequency (KHz) 2500 3000 −90 −100 0 0.1 0.2 0.3 0.4 Frequency Normalized To Filter Input Sample Rate Figure 22 Figure 24 0.5 The complete channel filter response for standard coefficients is shown in Figure 24. Passband flatness is shown in Figure 25. F2 STD frequency response A second set of F2 coefficients (GSM set, see Figure 23) suitable for meeting the stringent wideband GSM requirements with a gain of 0.999 are: The complete filter response for GSM coefficients is shown in Figure 26. GSM Passband flatness is shown in Figure 27. The mask shown in Figure 26 is derived from the ETSI GSM 5.05 specifications for a normal Basestation Transceiver (BTS). For interferers, 9dB was added to the carrier to interference (C/I) ratios. For blockers, 9dB was added to the difference between the blocker level and 3dB above the reference sensitivity level. -536, -986, 42, 962, 869, 225, 141, 93, -280, -708, -774, -579, -384, -79, 536, 1056, 1152, 1067, 789, 32, -935, -1668, -2104, -2137, -1444, 71, 2130, 4450, 6884, 9053, 10413, 10832 Rev. 3.05 May 27, 1999 CIC, F1, & F2 STD frequency response 16 ©1999 National Semiconductor Corporation Overall Channel Gain Combined Frequency Response of CIC/F1/F2 Using STD Set 0.1 The overall gain of the chip is a function of the amount of decimation (N), the settings of the “SHIFT UP” circuit (SCALE), the GAIN setting, the sum of the F1 coefficients, and the sum of the F2 coefficients. The overall gain is shown below in Equation 2. 0.08 0.06 Magnitude (dB) 0.04 0.02 0 1 4 G DDC = --- ( DEC + 1 ) 2 −0.02 −0.04 −0.06 −0.08 −0.1 0 Figure 25 50 100 150 Frequency (KHz) 200 ⋅2 [ SCALE – 44 – AGAIN ⋅ ( 1 – EXP_INH ) ] ⋅2 GAIN ⋅ G F1 ⋅ G F2 250 Where: CIC, F1, & F2 STD Passband Flatness 21 Combined Frequency Response of CIC/F1/F2 Using GSM Set ∑ h1( i ) 0 =1 G F1 = i--------------------16 2 FCK = 52MHz Decimation = 192 OFS = 270.83kHz −20 Magnitude (dB) EQ. 2 EQ. 3 and: −40 63 ∑ h2( i ) −60 =1 G F2 = i--------------------16 2 −80 It is assumed that the DDC output words are treated as fractional 2’s complement words. The numerators of G F1 −100 0 Figure 26 500 1000 1500 2000 Frequency (KHz) 2500 EQ. 4 3000 and G F2 equal the sums of the impulse response coefficients of F1 and F2, respectively. For the STD and GSM sets, G F1 and G F2 are nearly equal to unity. Observe that CIC, F1, & F2 GSM frequency response the AGAIN term in EQ. 2 is cancelled by the DVGA operation so that the entire gain of the DRCS is independent of Combined Frequency Response of CIC/F1/F2 Using GSM Set 1 1 the DVGA setting when EXP_INH=0. The --- appearing in 2 0.5 EQ. 2 is the result of the 6dB conversion loss in the mixer. Magnitude (dB) 0 1 For full-scale square wave inputs the --- should be set to 1 2 −0.5 to prevent signal distortion. Data Latency and Group Delay −1 The latency from a sync event to data output is approximately 8N CK periods for F2 decimation by 2 and 24N CK periods for F2 decimation by 4. Actual non-zero data output can be further delayed depending on the F1 and F2 filter coefficients. −1.5 −2 0 Figure 27 20 40 60 Frequency (KHz) 80 100 CIC, F1, & F2 GSM Passband Flatness ©1999 National Semiconductor Corporation Group Delay is approximately 90N CK periods for F2 decimation by 2 or 4. 17 Rev. 3.05 May 27, 1999 Serial Outputs Output Modes After processing by the DDC, the data is then formatted for output. Note All output data is two’s complement. RDY_POL, SCK_POL, SFS_POL MUX_MODE SERIAL OUTPUTS RDY MUX_MODE AOUT SCK SFS AOUT BOUT MUX POUT[15..0] 16 POUT_SEL[2..0] 3 POUT_EN Figure 28 CLC5902 output circuit The channel outputs are accessible through serial output pins and a 16-bit parallel output port. The RDY pin is provided to notify the user that a new output sample period (OSP) has begun. OSP refers to the interval between output samples at the decimated output rate. For example, if the input rate (and clock rate) is 52 MHz and the overall decimation factor is 192 (N=48, F2 decimation=2) the OSP will be 3.69 microseconds which corresponds to an output sample rate of 270.83kHz. An OSP starts when a sample is ready and stops when the next one is ready. BOUT 0 OUTA OUTB 1 OUTA, OUTB LOW Table 2 PACKED NUMBER FORMAT CONTROL SERIALIZER AND TDM FORMATTER DIVIDE BY RATE FORMAT CH A CH B CK POLARITY INVERT Output formats include truncation to 8 or 32 bits, rounding to 16 or 24 bits, and a 12-bit floating point format (4-bit exponent, 8-bit mantissa, 138dB numeric range). This function is performed in the OUTPUT CIRCUIT shown in Figure 28. The CLC5902 provides a serial clock (SCK), a frame strobe (SFS) and two data lines (AOUT and BOUT) to output serial data. The MUX_MODE control register specifies whether the two channel outputs are transmitted on two separate serial pins, or multiplexed onto one pin in a time division multiplexed (TDM) format. Separate output pins are not provided for the I and Q halves of complex data. The I and Q outputs are always multiplexed onto the same serial pin. The I-component is output first, followed by the Q-component. By setting the PACKED mode bit to ‘1’ a complex pair may be treated as a single double-wide word. The RDY signal is used to identify the first word of a complex pair of the TDM formatted output. The TDM modes are summarized in Table 2. TDM Modes The serial outputs use the format shown in Figure 29. Figure 29(a) shows the standard output mode (the PACKED mode bit is low). The chip clocks the frame and data out of the chip on the rising edge of SCK (or falling edge if the SCK_POL bit in the input control register is set high). Data should be captured on the falling edge of SCK (rising if SCK_POL=1). The chip sends the I data first by setting SFS high (or low if SFS_POL in the input control register is set high) for one clock cycle, and then transmitting the data, MSB first, on as many SCK cycles as are necessary. Without a pause, the Q data is transferred next as shown in Figure 29(a). If the PACKED control bit is high, then the I and Q components are sent as a double length word with only one SFS strobe as shown in Figure 29(b). If both channels are multiplexed out the same serial pin, then the subsequent I/Q channel words will be transmitted immediately following the first I/Q pair as shown in Figure 29(c). Figure 29(c) also shows how the RDY signal can be used to identify the I and Q channels in the TDM serial transmission.The serial output rate is programmable using the RATE register as a integer division of the input clock, the division ratio ranging from 1 to 256. The serial interface will not work properly if the programmed rate of SCK is insufficient to clock out all the bits in one OSP. Serial Port Output Number Formats Several numeric formats are selectable using the FORMAT control register. The I/Q samples can be rounded to 16 or 24 bits, or truncated to 8 bits. The packed mode works as described above for these fixed point formats. A floating point format with 138dB of dynamic range in 12 bits is also provided. The mantissa (m) is 8 bits and the exponent (e) is 4 bits. The MSB of each segment is trans- Note The serial outputs power up in a tri-state condition and must be enabled when the chip is configured. Parallel outputs are enabled by the POUT_EN pin. Rev. 3.05 May 27, 1999 18 ©1999 National Semiconductor Corporation clock stops and data is zero after transfers are complete SCK SFS AOUT I15 I1 I14 I0 Q15 Q14 Q1 Q0 (a) UNPACKED MODE, FRAME SYNC AT THE START OF EACH WORD clock stops and data is zero after transfers are complete SCK SFS I15 AOUT I14 I1 I0 Q15 Q14 Q1 Q0 (b) PACKED MODE, ONE FRAME SYNC AT THE START OF EACH DOUBLE-WORD TRANSFER RDY leading edge of RDY aligns with leading edge of SFS RDY is 4 CK periods wide Output Sample Period (OSP) SFS AOUT IA QA SFS AOUT IA QA IB QB MUX_MODE=0 IA QA MUX_MODE=1 IA QA IB QB (c) ONE OR TWO CHANNEL MUX MODES (PACKED MODE IS ON) clock stops and data is zero after transfers are complete SCK SFS AOUT mI7 mI6 mI0 eI3 eI2 eI0 eQ3 eQ0 mQ7 mQ0 (d) FLOATING POINT FORMAT Figure 29 Serial output formats. Refer to Figure 9 for detailed timing information mitted first. When this mode is selected, the I/Q samples are packed regardless of the state of MUX_MODE, and the data is sent as mI/eI/eQ/mQ which allows the two exponents to form an 8-bit word. This is shown in Figure 29(d). For all formats, once the defined length of the word is complete, SCK stops toggling. can be formatted as floating point numbers with an 8-bit mantissa and a 4 bit exponent. For the fixed-point formats, the valid bits are justified into the MSBs of the registers of Table 3 and all other bits are set to zero. For the floating POUT_SEL Parallel Outputs Output data from the channels can also be taken from a 16-b it par allel po rt. A 3-bit word applied to the POUT_SEL[2:0] pins determines which 16-bit segment is multiplexed to the parallel port. Table 3 defines this mapping. To allow for bussing of multiple chips, the parallel port is tri-stated unless POUT_EN is low. The RDY signal indicates the start of an OSP and that new data is ready at the parallel output. The user has one OSP to cycle through whichever registers are needed. The RATE register must be set so that each OSP is at least 5 SCK periods. The I/Q samples can be rounded to 16 or 24 bits or the full 32 bit word can be read. By setting the word size to 32 bits it is possible to read out the top 16 bits and only observe the top 8 bits if desired. Additionally, the output samples ©1999 National Semiconductor Corporation 19 Floating Point Register Contents 0 IA upper 16 bits 0000/eIA/mIA 1 IA lower 16 bits 0x0000 2 QA upper 16 bits 0000/eQA/mQA 3 QA lower 16 bits 0x0000 4 IB upper 16 bits 0000/eIB/mIB 5 IB lower 16 bits 0x0000 6 QB upper 16 bits 0000/eQB/mQB 7 QB lower 16 bits 0x0000 Table 3 Parallel Port Output Numeric Formats Normal Register Contents Register Selection for Parallel Output point format, the valid bits are placed in the upper 16 bits of the appropriate channel register using the format 0000/ eI/mI for the I samples. Rev. 3.05 May 27, 1999 AGC The CLC5902 AGC processor monitors the output level of the ADC and servos it to the desired setpoint. The ADC input is controlled by the DVGA to maintain the proper setpoint level. DVGA operation results in a compression of the signal through the ADC. The DVGA signal compression is reversed in the CLC5902 to provide > 120dB of linear dynamic range. This is illustrated in Figure 30. Diversity Receiver Chipset Full Scale Output Power A O GC ve O rT p hi era s R te an s ge D D C pu ut O 6dB steps the deadband should always be greater than 6dB to prevent oscillation. An increased deadband value will reduce the amount of AGC operation. A decreased deadband value will increase the amount of AGC operation but will hold the ADC output closer to the setpoint. The threshold should be set so that transients do not cause sustained overrange at the ADC inputs. The threshold setting can also be used to set the ADC input near its optimal performance level. The AGC may be configured to free run, operate for a programmable period of time (burst mode), or stop running. If a burst start pulse is applied to the AGC_EN pin in burst mode, the AGC will free run for AGC_HOLD_COUNT CK periods then freeze. At the beginning of the next burst either the freeze value is maintained or an initial condition is set from AGC_IC_A|B based on AGC_RESET_EN. Allowing the AGC to free run should be appropriate for most applications. Table 4 shows the AGC mode control bits. t ADC Full Scale AGC Threshold AGC Mode ADC Output Deadband+Hysteresis Figure 30 Input Power 6dB AgcRstEn AgcForce Free Run 0 0 1 Burst, hold previous value 0 0 0 Burst, use initial conditions 0 1 0 Manual 1 0 0 Table 4 Output Gain Scaling vs. Input Signal In order to use the AGC the DRCS Control Panel software should be used to calculate the programmable parameters. To generate these parameters only the desired setpoint, deadband+hysteresis, and loop time constant need to be supplied. All subsequent calculations are performed by the software. Complete details of the AGC operation are provided in an appendix but are not required reading. DVGA Output Power AgcHldIC AGC Operating Modes Power Management The CLC5902 can be placed in a low power (static) state by stopping the input clock. To prevent this from placing the CLC5902 into unexpected states, the SI pin of the CLC5902 should be asserted prior to disabling the input clock and held asserted until the input clock has returned to a stable condition. Deadband Reference Setpoint 6dB Hysteresis=Deadband-6dB DVGA Input Power Figure 31 AGC Setup. AGC setpoint and deadband+hysteresis are illustrated in Figure 31. The loop time constant is a measure of how fast the loop will track a changing signal. Values down to approximately 1.0 microsecond will be stable with the second order LC noise filter. Since the DVGA operates with Rev. 3.05 May 27, 1999 20 ©1999 National Semiconductor Corporation Real-time access to the following signals is provided by configuring the control interface debug register: Test and Diagnostics The CLC5902 supports IEEE 1149.1 compliant JTAG Boundary Scan for the I/O’s. The following pins are used: TRST TMS TDI TDO TCK (test reset) (test mode select) (test data in) (test data out) (test clock) • • • • NCO sine and cosine outputs data after round following mixers data before F1 and F2 data after the CIC filter within the AGC The access points are multiplexed to a 20-bit parallel output port which is created from signal pins POUT[15:0], AOUT, BOUT, SFS, and RDY according to the table below: The following JTAG instructions are supported: Normal Mode Pin Debug Mode Pin Instruction Description POUT[15:0] DEBUG[19..4] BYPASS Connects TDI directly to TDO RDY DEBUG[3] EXTEST Drives the ‘extest’ TAP controller output SFS DEBUG[2] IDCODE Connects the 32-bit ID register to TDO AOUT DEBUG[1] SAMPLE/PRELOAD Drives the ‘samp_load’ TAP controller output BOUT DEBUG[0] HIGHZ Tri-states the outputs The JTAG Boundary Scan can be used to verify printed circuit board continuity at the system level. The user is able to program a value into TEST_REG and substitute this for the normal channel inputs from the AIN/ BIN pins by selecting it with the crossbar. With the NCO frequency set to zero this allows the DDCs and the output interface of the chip to be verified. Also, the AGC loop can be opened by setting AGC_HOLD_IC high and setting the gain of the DVGA by programming the appropriate value into the AGC_IC_A/B register. SCK will be set to the proper strobe rate for each debug tap point. POUT_EN and PSEL[2..0] have no effect in Debug Mode. The outputs are turned on when the Debug Mode bit is set. Normal serial outputs are also disabled. Control Registers The chip is configured and controlled through the use of 8bit control registers. These registers are accessed for reading or writing using the control bus pins (CE, RD, WR, A[7:0], and D[7:0]) described in the Control Interface section. The register names and descriptions are listed in Table 5. Control Registers Register Name Width Type Defaulta Addr Bit Description DEC 11b R/W 7 0(LSBs) 1(MSBs) 7:0 2:0 CIC decimation control. N=DEC+1. Valid range is from 7 to 2047. Format is an unsigned integer. This affects both channels. DEC_BY_4 1b R/W 0 1 4 Controls the decimation factor in F2. 0=Decimate by 2. 1=Decimate by 4. This affects both channels. SCALE 6b R/W 0 2 5:0 CIC SCALE parameter. Format is an unsigned integer representing the number of left bit shifts to perform on the data prior to the CIC filter. Valid range is from 0 to 40. This affects both channels. GAIN_A 3b R/W 0 3 2:0 Value of left bit shift prior to F1 for channel A. GAIN_B 3b R/W 0 4 2:0 Value of left bit shift prior to F1 for channel B. RATE 1B R/W 1 5 7:0 Determines rate of serial output clock. The output rate is FCK/(RATE+1). Unsigned integer. SOUT_EN 1b R/W 0 6 0 Enables the serial output pins AOUT, BOUT, SCK, and SFS. 0=Tristate. 1=Enabled. SCK_POL 1b R/W 0 6 1 Determines polarity of the SCK output. 0=AOUT, BOUT, and SFS change on the rising edge of SCK (capture on falling edge). 1=They change on the falling edge of SCK. SFS_POL 1b R/W 0 6 2 Determines polarity of the SFS output. 0=Active High. 1=Active Low. RDY_POL 1b R/W 0 6 3 Determines polarity of the RDY output. 0=Active High. 1=Active Low. MUX_MODE 1b R/W 0 6 4 Determines the mode of the serial outputs. 0=Each channel is output on its respective pin, 1=Both channels are multiplexed and output on AOUT. See also Table 2. PACKED 1b R/W 0 6 5 Controls when SFS goes active. 0=SFS pulses prior to the start of the I and the Q words. 1=SFS pulses only once prior to the start of each I/Q sample pair (i.e. the pair is treated as a double-sized word) The I word precedes the Q word. See Figure 29. FORMAT 2b R/W 0 6 6:7 Determines output number format. 0=Truncate serial output to 8 bits. Parallel output is truncated to 32 bits. 1=Round both serial and parallel to 16 bits. All other bits are set to 0. 2=Round both serial and parallel to 24 bits. All other bits are set to 0. 3=Output floating point. 8-bit mantissa, 4-bit exponent. All other bits are set to 0. Table 5 CLC5902 Control Registers ©1999 National Semiconductor Corporation 21 Rev. 3.05 May 27, 1999 Register Name Width Type Defaulta Addr Bit Description FREQ_A 4B R/W 0 7-10 7:0 Frequency word for channel A. Format is a 32-bit, 2’s complement number spread across 4 registers. The LSBs are in the lower registers. The NCO frequency F is F/FCK=FREQ_A/232. PHASE_A 2B R/W 0 11-12 7:0 Phase word for channel A. Format is a 16-bit, unsigned magnitude number spread across 2 registers. The LSBs are in the lower registers. The NCO phase PHI is PHI=2*pi*PHASE_A/ 2^16. FREQ_B 4B R/W 0 13-16 7:0 Frequency word for channel B. Format is a 32-bit, 2’s complement number spread across 4 registers. The LSBs are in the lower registers. The NCO frequency F is F/FCK=FREQ_B/232. PHASE_B 2B R/W 0 17-18 7:0 Phase word for channel B. Format is a 16-bit, unsigned magnitude number spread across 2 registers. The LSBs are in the lower registers. The NCO phase PHI is PHI=2*pi*PHASE_B/ 2^16. A_SOURCE 2 R/W 0 19 1:0 0=Select AIN as channel input source. 1=Select BIN. 2=3=Select TEST_REG as channel input source. B_SOURCE 2 R/W 1 19 2:3 0=Select AIN as channel input source. 1=Select BIN. 2=3=Select TEST_REG as channel input source. EXP_INH 1b R/W 0 20 0 0=Allow exponent to pass into FLOAT TO FIXED converter. 1=Force exponent in DDC channel to a 7 (maximum digital gain). This affects both channels. AGC_FORCE 1b R/W 1 20 1 0=Enable AGC counter operation. 1=AGC loop operates continuously regardless of AGC_EN pin. This affects both channels. AGC_RESET_EN 1b R/W 0 20 2 0=Initial condition is never used. 1=Integrator is reset each time the AGC transitions from idle to active. This affects both channels. AGC_HOLD_IC 1b R/W 0 20 3 0=Normal closed-loop operation. 1=Hold integrator at initial condition. This affects both channels. AGC_LOOP_GAIN 2b R/W 0 20 4:5 Bit shift value for AGC loop. Valid range is from 0 to 3. This affects both channels. AGC_COUNT 2B R/W 0 21-22 7:0 Counter value for AGC enable counter. Format is a 16-bit, unsigned magnitude number spread across 2 registers. The LSBs are in the lower register. The value represents the number of CK cycles over which the loop is active. AGC_IC_A 1B R/W 0 23 7:0 AGC integrator initial condition for channel A. Format is an 8-bit, unsigned magnitude number. This number is written into the magnitude MSBs of the channel A AGC integrator whenever it is reset to the initial condition. AGC_IC_B 1B R/W 0 24 7:0 AGC integrator initial condition for channel B. Format is an 8-bit, unsigned magnitude number. This number is written into the magnitude MSBs of the channel B AGC integrator whenever it is reset to the initial condition. AGC_RB_A 1B R 0 25 7:0 AGC integrator readback value for channel A. Format is an 8-bit, unsigned magnitude number. The user can read the magnitude MSBs of the channel A integrator from this register. AGC_RB_B 1B R 0 26 7:0 AGC integrator readback value for channel B. Format is an 8-bit, unsigned magnitude number. The user can read the magnitude MSBs of the channel B integrator from this register. TEST_REG 14b R/W 0 27(LSBs) 28(MSBs) 7:0 5:0 Test input source. Instead of processing values from the AIN/BIN pins, the value from this location is used instead. Format is 14-bit 2s complement number spread across 2 registers. Reserved 1B - - 29 7:0 For future use. Reserved 1B - - 30 7:0 For future use. DEBUG_EN 1b R/W 0 31 0 0=Normal. 1=Enables access to the internal probe points. DEBUG_TAP 5b R/W 0 31 1:5 Selects internal node tap for debug. 0 selects F1 output for BI, 20 bits 1 selects F1 output for BQ, 20 bits 2 selects F1 output for AQ, 20 bits 3 selects F1 output for AI, 20 bits 4 selects F1 input for BI, 20 bits 5 selects F1 input for BQ, 20 bits 6 selects F1 input for AI, 20 bits 7 selects F1 input for AQ, 20 bits 8 selects NCO A, cosine output. 17 bits, 3 LSBs are 0. 9 selects NCO A, sine output, 17 bits, 3 LSBs are 0. 10 selects NCO B, cosine output, 17 bits, 3 LSBs are 0. 11 selects NCO B, sine output, 17 bits, 3 LSBs are 0. 12 selects NCO AI, rounded output, 15 bits, 5 LSBs are 0. 13 selects NCO AQ, rounded output, 15 bits, 5 LSBs are 0. 14 selects NCO BI, rounded output, 15 bits, 5 LSBs are 0. 15 selects NCO BQ, rounded output, 15 bits, 5 LSBs are 0. 16-31 selects AGC CIC filter output. 9 MSBs from ch A, next 9 bits from ch B, 2 LSBs are 0. DITH_A 1b R/W 1 31 6 0=Disable NCO dither source for channel A. 1=Enable. DITH_B 1b R/W 1 31 7 0=Disable NCO dither source for channel B. 1=Enable. AGC_TABLE 32B R/W 0 128-159 7:0 RAM space that defines key AGC loop parameters. Format is 32 separate 8-bit, 2’s complement numbers. This is common to both channels. F1_COEFF 22B R/W 0 160-181 7:0 Coefficients for F1. Format is 11 separate 16-bit, 2’s complement numbers, each one spread across 2 registers. The LSBs are in the lower registers. For example, coefficient h0[7:0] is in address 160, h0[15:8] is in address 161, h1[7:0] is in address 162, h1[15:8] is in address 163. F2_COEFF 64B R/W 0 182-245 7:0 Coefficients for F2. Format is 32 separate 16-bit, 2’s complement numbers, each one spread across 2 registers. The LSBs are in the lower registers. For example, coefficient h0[7:0] is in address 182, h0[15:8] is in address 183, h1[7:0] is in address 184, h1[15:8] is in address 185. Table 5 CLC5902 Control Registers a. These are the default values set by a master reset (MR). Sync in (SI) will not affect any of these values. Rev. 3.05 May 27, 1999 22 ©1999 National Semiconductor Corporation Condensed Hexadecimal Address Map Reg Nam e DEC D EC _BY_4 SCALE G A IN 1 _ A G A IN 1 _ B RATE S E R IA L _ C T R L FR EQ _A PH ASE_A FR EQ _B PH ASE_B SOURCE AG C _C TR L AG C _C O U N T A G C _ IC _ A A G C _ IC _ B AG C _R B_A AG C _R B_A TEST_R EG DEBUG AG C _TABLE F1_C O EFF F2_C O EFF Addr [ [ [ [ [ [ [ [ [ [ [$ [% [& [' [( [) [ [ [ [ [ [ [ [ [ [ [$ [% [& [) [ [) [$ [% [% [) B it7 D ec7 B it6 D ec6 R a te 7 FM T1 FA7 FA15 FA23 FA31 PA7 PA15 FB7 FB15 FB23 FB31 PB7 PB15 R a te 6 FM T0 FA6 FA14 FA22 FA30 PA6 PA14 FB6 FB14 FB22 FB30 PB6 PB14 A g c C n t7 A g c C n t1 5 A g c Ic A 7 A g c Ic B 7 AgcR bA 7 AgcR bB 7 T e s t7 A g c C n t6 A g c C n t1 4 A g c Ic A 6 A g c Ic B 6 A gcR bA 6 A gcR bB 6 T e s t6 D IT H _ B D IT H _ A ©1999 National Semiconductor Corporation B it5 D ec5 S c a le 5 B it 4 D ec4 D e cB y4 S c a le 4 R a te 5 Packed FA5 FA13 FA21 FA29 PA5 PA13 FB5 FB13 FB21 FB29 PB5 PB13 R a te 4 M uxM ode FA4 FA12 FA20 FA28 PA4 PA12 FB4 FB12 FB20 FB28 PB4 PB12 AgcLG 1 A g c C n t5 A g c C n t1 3 A g c Ic A 5 A g c Ic B 5 AgcR bA5 AgcR bB5 T e s t5 T e s t1 3 T a p S e l4 AgcLG 0 A g c C n t4 A g c C n t1 2 A g c Ic A 4 A g c Ic B 4 AgcR bA4 AgcR bB4 T e s t4 T e s t1 2 T a p S e l3 B it 3 D ec3 B it 2 D ec2 D ec10 S c a le 3 S c a le 2 G 1A2 G 1B2 R a te 3 R a te 2 R D Y_PO L SFS_PO L FA3 FA2 FA11 FA10 FA19 FA18 FA27 FA26 PA3 PA2 PA11 PA10 FB3 FB2 FB11 FB10 FB19 FB18 FB27 FB26 PB3 PB2 PB11 PB10 B S rc 1 B S rc 0 A g c H ld IC A g c R s t E n A g c C n t3 A g c C n t2 A g c C n t1 1 A g c C n t 1 0 A g c Ic A 3 A g c Ic A 2 A g c Ic B 3 A g c Ic B 2 AgcR bA3 AgcR bA2 AgcR bB3 AgcR bB2 T e s t3 T e s t2 T e s t1 1 T e s t1 0 T a p S e l2 T a p S e l1 B it 1 D ec1 D ec9 S c a le 1 G 1A1 G 1B1 R a te 1 SCK_PO L FA1 FA9 FA17 FA25 PA1 PA9 FB1 FB9 FB17 FB25 PB1 PB9 A S rc 1 A g c F o rce A g c C n t1 A g c C n t9 A g c Ic A 1 A g c Ic B 1 AgcR bA1 AgcR bB1 T e s t1 T e s t9 T a p S e l0 B it 0 D ec0 D ec8 S c a le 0 G 1A0 G 1B0 R a te 0 SO U T_EN FA0 FA8 FA16 FA24 PA0 PA8 FB0 FB8 FB16 FB24 PB0 PB8 A S rc 0 E x p In h A g c C n t0 A g c C n t8 A g c Ic A 0 A g c Ic B 0 AgcR bA0 AgcR bB0 T e s t0 T e s t8 D ebugEn T h e A g c T a b le lo a d s f r o m t h e lo w a d d r e s s to t h e h ig h a d d r e s s in t h is o r d e r : 1 s t lo c a t io n , 2 n d lo c a t io n … T h e F I R C o e f f ic ie n t s lo a d f r o m t h e lo w a d d r e s s t o t h e h ig h a d d r e s s in t h is o r d e r 1 s t lo c a t io n lo w b y t e , 1 s t lo c a t io n h ig h b y t e , 2 n d lo c a t io n … 23 Rev. 3.05 May 27, 1999 The power detector bandwidth is set by the CIC filter to FCK/8. The absolute value circuit doubles the effective input frequency. This has the effect of reducing the power detector bandwidth from FCK/8 to FCK/16. AGC Theory of Operation A block diagram of the AGC is shown in Figure 32. The DVGA interface comprises four pins for each of the channels. The first three pins of this interface are a 3-bit binary word that controls the DVGA gain in 6dB steps (AGAIN). The final pin is ASTROBE which allows the AGAIN bits to be latched into the DVGA’s register. A key feature of the ASTROBE, illustrated Figure 33, is that it toggles only if the data on AGAIN has changed from the previous cycle. Not shown is that ASTROBE and BSTROBE are independent. For example, ASTROBE should only toggle when AGAIN has changed. BSTROBE should not toggle because AGAIN has changed. This is done to minimize unnecessary digital noise on the sensitive analog path through the DVGA. ASTROBE and BSTROBE are asserted during MR and SI to properly initialize the DVGAs. For a full-scale sinusoidal input, the absolute value circuit output is a dc value of 511 ⋅ ( 2 ⁄ π ) . Because the absolute value circuit also generates undesired even harmonic terms, the CIC filter (response shown in Figure 34), is required to remove these harmonics. The first response null occurs at FCK/8, where FCK is the clock frequency, and the response magnitude is at least 25dB below the dc value from FCK/10 to 9FCK/10. Because the 2nd harmonic from the absolute value circuit is about 10dB below the dc this means that the ripple in the detected level is about 0.7dB or less for input frequencies between F CK /20 to 19FCK/20. The “FIXED TO FLOAT CONVERTER” takes the fixed point 9-bit output from the CIC filter and converts it to a “floating point” number. This conversion is done so that the 32 values in the RAM can be uniformly assigned (dB The absolute value circuit and the 2-stage, decimate-by-8 CIC filter comprise the power detection part of the AGC. AGC_TABLE AGC_LOOP_GAIN 8 12 AGAIN 3 12 MUX 5 SHIFT DOWN 9 32X8 RAM POUT FIXED TO FLOAT CONVERTER 16 2 STAGE DECIMATE BY 8 CIC FILTER 10 SHIFT DOWN AIN[13:4] ABSOLUTE VALUE EXP AGC_IC_A AGC_HOLD_IC AGC_RESET_EN AGC_FORCE LOAD ONE SHOT AGC_EN PERIOD COUNTER AGC_COUNT 16 FUNCTION PROGRAMMED INTO RAM LOG EN -REF Figure 32 CLC5902 AGC circuit, Channel A ASTROBE does not pulse because AGAIN[2:0] does not change 1 SCK delay CK CK/8 ASTROBE AGAIN[2:0] Figure 33 Timing diagram for AGC/DVGA interface, Channel A. Refer to Figure 8 for detailed timing information. Rev. 3.05 May 27, 1999 24 ©1999 National Semiconductor Corporation As shown in Figure 32, the 32X8 RAM look-up table implements the functions of log converter, reference subtraction, error amplifier, and deadband. The user must build each of these functions by constructing a set of 8-bit, 2’s complement numbers to be loaded into the RAM. Each of these functions and how to construct them are discussed in the following paragraphs. 0 Magnitude (dB) -10 -20 -30 A log conversion is done in order to keep the loop gain independent of operating point. To see why this is beneficial, the control gain of the DVGA computed without log conversion is, -40 -50 ′ -60 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 K DVGA = 1 Frequency Normalized to Sample Rate Figure 34 = – v i ⋅ ln ( 2 ) ⋅ 2 Power detector filter response scale) to detected power levels (54 db range). This provides a resolution of 1.7dB between detected power levels. The truth table for this converter is given in Table 6. The upper three bits of the output represent the exponent (e) OUTPUT (eeemm) INPUT 0-3 000XX 4-7 001XX 8-15 010XX 16-31 011XX 32-63 100XX 64-127 101XX 128-255 110XX 256-511 111XX Table 6 2 , where G is the decimal equivalent of GAIN and G o accounts for the DVGA gain in excess of unity. This equation assumes that the DVGA gain control polarity is positive as is the case for the CLC5526. The gain around the entire loop must be negative. Observe in Equation 6 that the control gain is dependent on operating point G. If we instead compute the control gain with log conversion, EQ. 7 which is no longer operating-point dependent. The log function is constructed by computing the CIC filter output associated with each address (Equation 5) and converting these to dB. Full scale (dc signal) is 20 log ( 511 ) = 54dB . EQ. 5 , E ≥ 1. The max() and min() operators account for row 1 of Table 6 which is a special case because M=POUT. Equation 5 associates each address of the RAM with a CIC filter output. ©1999 National Semiconductor Corporation EQ. 6 = – 6.02, and the lower 2 are the mantissa (m). The exponent is determined by the position of the leading ‘1’ out of the CIC filter. An output of ‘001XX’ corresponds to a leading ‘1’ in bit 2 (LSB is bit 0). The exponent increases by one each time the leading ‘1’ advances in bit position. The mantissa bits are the two bits that follow the leading ‘1’. If we define E as the decimal value of the exponent bits and M as the decimal value of the mantissa bits, the output of the CIC filter, POUT, corresponding to a given “FIXED TO FLOAT CONVERTER” output is, ( max ( E, 1 ) – 1 ) ( G – Go) ( G – Go) K DVGA = ∂ [ 20 ⋅ log ( v i ⋅ 2 ) ], ∂G Fixed to Float Converter Truth Table P OUT = [ 4 ⋅ min ( E, 1 ) + M ] ⋅ ( G – Go) ∂ (v ⋅ 2 ), ∂G i 25 The reference subtraction is constructed by subtracting the desired loop servo point (in dB) from the table values computed in the previous paragraph. For example, if it is desired that the DVGA servo the ADC input level (sinusoidal signal) to -6dBFS, the number to subtract from the data is 511 2 20 log --------- ⋅ --- = 44dB . 2 π EQ. 8 The table data will then cross through zero at the address corresponding to this reference level. A deadband wider than 6dB should then be constructed symmetrically about this point. This prevents the loop from hunting due to the 6dB gain steps of the DVGA. Any deadband in excess of 6dB appears as hysteresis in the servo point of the loop as illustrated in Figure 31. The deadband is constructed by loading zeros into those addresses on either side of the one which corresponds to the reference level. The last function of the RAM table is that of error amplification. All the operations preceding this one gave a table slope S RAM = 1 . This must now be adjusted in order to control the time constant of the loop given by, Rev. 3.05 May 27, 1999 20 AGC RAM CONTENTS AGC RAM CONTENTS 20 0 -20 -40 -60 -80 0 5 10 -40 -60 -80 15 20 25 -120 30 0 10 20 30 40 ADDRESS POUT (dB) (a) (b) 50 60 Example of programmed RAM contents 8 1- + 1--- . τ = ---------- -----F CK G L 2 EQ. 9 The term GL in this equation is the loop gain, G L = – 6.02 ⋅ S RAM ⋅ 2 ( AGC_LOOP_GAIN – 4 ) . EQ. 10 The design equations are obtained by solving Equation 9 for GL and Equation 10 for S RAM . AGC_LOOP_GAIN is a control register value that determines the number of bits to shift the output of the RAM down by. This allows some of the loop gain to be moved out of the RAM so that the full output range of the table is utilized but not exceeded. The valid range for AGC_LOOP_GAIN is from 0 to 3 which corresponds to a 1 to 4 bit shift left. An example set of numbers to implement a loop having a reference of 6dB below full scale, a deadband of 8dB, and a loop gain of 0.108 is: AGC_RB_A (or AGC_RB_B) register. The top 3 bits below the sign become AGAIN and are output along with ASTROBE signal on the DVGA interface pins. The valid range of AGAIN is from 0 to 7 which corresponds to a valid range of 0 to 210-1 for the 11-bit, 2’s complement integrator output from which AGAIN is derived. This is illustrated in Figure 36. The integrator saturates at these 7 For this range to be the same size as all others, the max integrator output must be limited to 8x27-1 =210-1 6 5 AGAIN Figure 35 -20 -100 -100 -120 0 4 3 2 -102 -102 -88 -80 -75 -70 -66 The min integrator output must be limited to 0 so that the sign of AGAIN is positive 1 -63 -61 -56 -53 -50 -47 -42 -39 -36 -33 -29 -25 0 0 -22 -19 -15 -11 0 0 0 0 0 13 17 2x27 3x27 4x27 5x27 6x27 7x27 8x27 Integrator Output 20 These values are shown plotted in Figure 35 with respect to the table addresses in (a), and the CIC filter output POUT in (b). For a 52MHz clock rate and AGC_LOOP_GAIN=2, these values result in a loop time constant of 1.5µs . The error signal from the loop gain “SHIFT DOWN” circuit is gated into the loop integrator. The gate is controlled by a timing and control circuit discussed in the next paragraph. A MUX within the integrator feedback allows the integrator to be initialized to the value loaded into AGC_IC_A (channel B can be set independently). The conditions under which it is initialized are configured in the registers associated with the timing and control circuit. The top eight bits of the integrator output can also be read back over the microprocessor interface from the Rev. 3.05 May 27, 1999 1x27 0 26 Figure 36 AGC integrator output limits limits to prevent overshoots as the integrator attempts to enter the valid range. The AGAIN value is inverted (EXP) and used to adjust the gain of the incoming signal to provide a linear output dynamic range. The relationship between the DVGA analog gain (AGAIN) and the “FIXED TO FLOAT CONVERTER” digital gain (EXP) is shown in Table 7. The DVGA’s compression of the incoming signal in the analog domain vs. the subsequent expansion in the digital domain is shown in Figure 30. Several control bits allow the user to configure a variety of AGC algorithms. The AGC may free run by setting AGC_FORCE high and AGC_HOLD_IC low. If a burst start pulse is available and sent to the AGC_EN pin, the ©1999 National Semiconductor Corporation AGAINa EXPb 000 = -12dB 111 = +0dB 001 = -6dB 110 = -6dB 010 = +0dB 101 = -12dB -12dB 14 14 14 13 12 11 011 = +6dB 100 = -18dB -12dB 14 14 14 14 13 12 11 100 = +12dB 011 = -24dB -12dB 14 14 14 14 14 13 12 11 ... 5 4 3 2 1 0 L L L 101 = +18dB 010 = -30dB -12dB 14 14 14 14 14 14 13 12 ... 6 5 4 3 2 1 0 L L 110 = +24dB 001 = -36dB -12dB 14 14 14 14 14 14 14 13 ... 7 6 5 4 3 2 1 0 L 111 = +30dB 000 = -42dB -12dB 14 14 14 14 14 14 14 14 ... 8 7 6 5 4 3 2 1 0 Table 7 Inputc 21 20 19 18 17 16 15 14 ... 8 7 6 5 4 3 2 1 0 -12dB 14 13 12 11 -12dB 14 14 13 12 11 10 9 8 7 ... 1 0 L L L L L L L 10 9 8 ... 2 1 0 L L L L L L 10 9 ... 3 2 1 0 L L L L L 10 ... 4 3 2 1 0 L L L L 15-bit Mixer Output Alignment into the 22-bit SHIFT-UP Based On EXP. d a. AGAIN sets the DVGA or analog gain value. b. EXP sets the “FIXED TO FLOAT CONVERTER” or digital gain value. c. 22-bit input to SHIFT-UP block in Figure 15 horizontally, linearized SHIFT-UP value vertically. d. The numbers in the center of the table represent the mixer output bits. ‘L’ represents a logic low. AGC may be configured to adapt the gain during the power ramp sequence then hold it steady afterwards by setting AGC_FORCE low, AGC_HOLD_IC low and programming AGC_HOLD_COUNT for the period of the power ramp. In addition, one can provide an initial condition and set AGC_RESET_EN high to have the gain start at a prescribed value at the beginning of each burst. The AGC becomes active when the AGC_EN pin goes low and remains active for AGC_HOLD_COUNT samples after AGC_EN goes high. The AGC_HOLD_COUNT can be programmed to one value during the random access burst and a smaller value during a normal burst. Finally, one might program a more narrow AGC loop bandwidth during normal transmission. Just prior to the next burst transmission from a mobile, set the initial condition to the value read back from the AGC accumulator at the end of the previous burst from that specific mobile. Allowing the AGC to free run should be appropriate for most applications. If the INH_EXP bit is not set, the DVGA gain word (EXP) is routed to the “FLOAT TO FIXED CONVERTER” in the DDCs prior to the programmable decimation filter. The EXP signals are delayed to account for the propagation delay of the DVGA interface and the CLC5956 ADC. The basic AGC modes are summarized in Table 4. ©1999 National Semiconductor Corporation 27 Evaluation Hardware Evaluation boards are available to facilitate designs based on the CLC5902: CLC-DRCS-PCASM The Diversity Receiver Chipset evaluation board providing a complete narrowband receiver from IF to digital symbols. CLC-CAPT-PCASM A simple method for capturing output data from CLC ADCs and the CLC5902. SOFTWARE Control panel software for the CLC5902 supports complete device configuration on both evaluation boards. Capture software manages the capture of data and its storage in a file on a PC. Matlab script files support data analysis: FFT, DNL, and INL plotting. This software and additional application information is available on the CLC Evaluation Kit CDROM. Rev. 3.05 May 27, 1999 CLC5902 Dual Digital Tuner/AGC Customer Design Applications Support National Semiconductor is committed to design excellence. For sales, literature and technical support, call the National Semiconductor Customer Response Group at 1-800-272-9959 or fax 1-800-737-7018. Life Support Policy National’s products are not authorized for use as critical components in life support devices or systems without the express written approval of the president of National Semiconductor Corporation. As used herein: 1. Life support devices or systems are devices or systems which, a) are intended for surgical implant into the body, or b) support or sustain life, and whose failure to perform, when properly used in accordance with instructions for use provided in the labeling, can be reasonably expected to result in a significant injury to the user. 2. A critical component is any component of a life support device or system whose failure to perform can be reasonably expected to cause the failure of the life support device or system, or to affect its safety or effectiveness. N National Semiconductor Corporation National Semiconductor Europe National Semiconductor Hong Kong Ltd. National Semiconductor Japan Ltd. 1111 West Bardin Road Arlington, TX 76017 Tel: 1(800) 272-9959 Fax: 1(800) 737-7018 Fax: (+49) 0-180-530 85 86 E-mail: europe.support.nsc.com Deutsch Tel: (+49) 0-180-530 85 85 English Tel: (+49) 0-180-532 78 32 Francais Tel: (+49) 0-180-532 93 58 Italiano Tel: (+49) 0-180-534 16 80 13th Floor, Straight Block Ocean Centre, 5 Canton Road Tsimshatsui, Kowloon Hong Kong Tel: (852) 2737-1600 Fax: (852) 2736-9960 Tel: 81-043-299-2309 Fax: 81-043-299-2408 National does not assume any responsibility for use of any circuitry described, no circuit patent licenses are implied and National reserves the right at any time without notice to change said circuitry and specifications. National Semiconductor® and N® are trademarks of National Semiconductor Corporation. All other trademarks are the property of their respective companies. http://www.national.com 28 Rev. 3.05 May 27, 1999