AN278 VOICE R ECORDER R EFERENCE D ESIGN 1. Introduction The C8051F411 offers a versatile, small (5 x 5 mm), highly integrated, low-power solution for voice applications. The 12-bit ADC and DAC allow for reasonable quality sound at a 8 kHz sampling rate, and the hardware Accumulation and Burst Mode features of the ADC provide for further improvements with small processing tradeoffs. The Suspend mode operating feature allows the voice recorder to "sleep" while idle, saving power in a similar fashion to the traditional 8051 Stop mode, but still allows the recorder to wake and respond to the user without a hardware reset. This document describes the solution for a telephone-quality voice recorder using the C8051F411. This document includes the following: A description of the system hardware and software Usage notes and customization considerations A schematic, bill of materials, and detailed layout diagrams The implementation of the software showing how to sample, compress, store, and play back a voice signal The code accompanying this application note was originally written for C8051F41x devices. It can also be ported to other devices in the Silicon Labs microcontroller range. 1.1. Key Points Because of its small size, versatile peripherals, and low-power features, the 'F411 readily lends itself to batteryoperated voice applications. The system uses a DPCM (Differential Pulse Code Modulation) compression algorithm for data storage to extend the total recording time. The recorder takes steps to minimize power usage while active and uses the "Suspend" feature of the 'F411 to reduce power consumption when idle. 2. Overview The system, depicted in Figure 1, consists of a microphone and speaker, input and output filtering, the microcontroller, the external Flash memory storage, and push-buttons and LEDs for user interaction. This section describes each aspect of this system. Push-buttons and LEDs Microphone Anti-aliasing filter and amplifier Filter and speaker driver Speaker ADC SST 512 k Flash Memory DAC C8051F411-GM Figure 1. Voice Recorder Physical System Overview Rev. 0.2 7/13 Copyright © 2013 by Silicon Laboratories AN278 AN278 2.1. Anti-Aliasing and Output Filtering Both the input to the ADC and the output from the IDAC are filtered through low-pass, op-amp filters. The filters before the ADC help prevent aliasing, where sound waveforms with frequencies above half of the sampling frequency (the Nyquist frequency) "sound" much lower because the sampling is not adequately fast enough to properly reconstruct the original waveform. The filtering on the output eliminates the high-frequency content of the IDAC output and smoothes out the waveform before it is passed through the speaker driver. 2.2. Microcontroller (C8051F411) The 'F411 samples the voice signal using the ADC, compresses the sample using DPCM (Differential Pulse Code Modulation), and sends the sample to the external Flash using the SPI. The microcontroller later retrieves the samples from the external Flash, decompresses them, and sends them to the speaker through the DAC. Figure 2 shows this dataflow path. ADC DPCM algorithm External Flash DPCM algorithm DAC Figure 2. Voice Recorder Dataflow Path To achieve telephone-quality sound, the microcontroller uses a sampling frequency of 8 kHz. This sampling rate can adequately reconstruct voice frequencies below 4 kHz, and still allots plenty of time between samples for the microcontroller to compress each sample and send it to external Flash memory. The microcontroller uses the hardware Burst Mode and Repeat Count features to automatically oversample and average the ADC samples, providing greater ADC accuracy. When the voice recorder is idle, the microcontroller shuts the system down using the Suspend mode feature, which minimizes the power consumption and allows the microcontroller to wake when the user presses either the Record/Play or Erase button without a hardware reset. 2.3. SPI Flash storage With DPCM, the 'F411 compresses each 12-bit ADC sample into 6 bits, so four samples can be stored in every 3 bytes. With an 8 kHz sampling rate and 32 kB of internal Flash, the 'F411 can store approximately 5-6 seconds' worth of recordings by itself. An additional 512 kB serial (SPI) Flash memory is included on the board to extend the total storage to 1 minute 27 seconds. 2.4. Push-buttons and LEDs The voice recorder uses simple LEDs and push-buttons to interact with the user. These LEDs indicate which function the recorder is using and whether the recorder is active or idle. In order to have the basic functionality of a voice recorder, the user needs to be able to record, play, erase, increase volume, and decrease volume. The two switches and potentiometer on the board provide these functions. 3. Hardware Description This section includes the detailed descriptions of the hardware components for the voice recorder. 3.1. Audio Paths The voice recorder includes an on-board microphone for mobility and ease of design, since different microphones require different biasing circuits. The microphone signal is sent through a rough low-pass filter and gain op-amp circuit to utilize the full range of the ADC, and then through a 3-pole Butterworth filter with a corner frequency of 4 kHz. The op-amps operate using 3.3 V rail-to-rail, but the ADC uses the programmed internal VREF of 2.2 V, so a voltage divider and DC-blocking capacitor provide the voltage translation from the filters to the ADC. A 5-pole Butterworth filter smoothes the output of the IDAC, which is then used by the speaker driver to output the waveform to the speaker jack. 2 Rev. 0.2 AN278 3.2. Low-power Suspend The 'F411 has a low power Suspend mode, during which the internal oscillator is completely dormant. An external transistor, controlled by one of the Port I/O, allows the 'F411 to disconnect the power to all of the external circuitry (op-amps, SST Flash, and speaker driver). Only the external voltage regulator and the 'F411 consume power while the system is idle. 3.3. MCU Peripherals The voice recorder uses four of the 'F411 peripherals: Analog-Digital Converter (ADC), Serial Peripheral Interface (SPI), Current Digital-Analog Converter (IDAC), and Programmable Counter Array (PCA). The 12-bit ADC samples the voice input and provides hardware accumulation and oversampling. The SPI communicates with the external Flash memory in 4-wire master-mode to store compressed samples. The 12-bit DAC outputs the decompressed sample from memory to the speaker driver. Finally, the PCA in 8-bit PWM mode controls the brightness of the LEDs for user interaction. 3.4. Layout Considerations This project does not include any extremely sensitive analog devices, so the main concerns during layout are size and cost. However, some care needs to be taken when routing peripherals and signals to the microcontroller Port I/O. For example, coupling can occur between the high-frequency SPI and the sensitive analog ADC and DAC peripherals, so these signals should be separated. Additionally, the DAC and VREF are only available on specific Port I/O pins. Furthermore, the SPI has higher precedence in the crossbar priority than the PCA when both are enabled, so the crossbar will route the SPI first. Careful planning of all I/O will ensure that all pins are routed correctly. 4. Software Description The voice recorder microcontroller is responsible for checking the switches for user interaction, sampling the voice input, compressing and decompressing the samples, storing the samples in external Flash, outputting the samples to the output filters, and controlling the PWM of the LEDs. This section describes each of these functions and their implementation in detail. 4.1. Push-buttons The 'F411 has two external interrupt pins that may be routed to any Port 0 pin. The voice recorder could successfully use these interrupts for the push-buttons, but this would limit the voice recorder design to only having two buttons. If the voice recorder is integrated into another design or if more features are added, more than two buttons would be needed. Instead, the voice recorder uses a polling scheme, where a Timer checks the switches periodically, but the rate at which they are pressed is relatively slow compared to the other functions the voice recorder performs. The switches need to be checked quickly enough that they're adequately responsive to the user, but not so quickly that they constantly toggle before the switch is released. To account for both needs, the switches are checked every 15 ms and a 150 ms delay is added every time a switch is pressed. 4.2. Sampling Considerations The sampling frequency for both the ADC and the DAC must be controlled as precisely as possible so that the output doesn't shift frequencies from the original input. The 'F411 uses the two 16-bit timers (Timer 2 and Timer 3) with auto reload to accomplish this. The ISR associated with each timer must be short enough that it doesn't interfere with the sampling frequency of 8 kHz, so each ISR execution must be less than 125 µs. Thus, all extraneous activities, such as the switch polling and LED control, must be executed in another, lower priority ISR that can be interrupted as necessary to meet the timing requirements of the sampling. Since these routines must communicate with one another, the 'F411 demo software uses global flags to indicate whether the sampling ISR should start, stop, or complete some other action. Rev. 0.2 3 AN278 4.3. DPCM Compression The voice recorder uses a DPCM, or Differential Pulse Code Modulation, compression scheme, which is lossy because of the error incurred due to the nature of the algorithm. This scheme reduces each 12-bit sample down to a 6-bit code representing the difference between the actual sample and the predicted value of the sample. This predicted value can be calculated from sample averaging or some other complex algorithm, but, because voice samples tend to be highly correlated and the ISR needs to be short, the voice recorder simply uses the previous iteration's result as the predictor. The DPCM compression algorithm is shown in Figure 3. Sn+1 Quantize to storage (External Flash) Encode Sn Pn Store predictor for one cycle Decode Pn+1 Figure 3. DPCM Compression Algorithm The difference between the predictor and the sample is quantized, or separated into different "ranges" or "bins," and the 6-bit code represents the 64 possible ranges of values. This coded and quantized difference is then stored in memory. To calculate the new predicted value, the compression algorithm then decodes the difference and adds it to the current predicted value. The decompression algorithm, shown in Figure 4, simply consists of matching the code in memory with the quantized difference and adding that difference to the predictor. to storage (External Flash) Sn+1 Decode Sn / Pn Store predictor for one cycle Pn+1 Figure 4. DPCM Decompression Algorithm The initial predicted values from both the compression and decompression schemes should match so that the DPCM input and output are as similar as possible. The following example uses a 4-bit (with 16 values) DPCM algorithm for an 8-bit ADC. If the first ADC sample is 0x89 (137 in unsigned decimal) and the initial predicted value is 0x80 (128 in unsigned decimal), the difference between the sample and the predicted value (sample - predicted value) is 9. When quantized, this difference of 9 falls in the "between 8 and 16" range, so the resulting DPCM code is 12. This code of 12 is then sent to storage. Difference DPCM code -64 1 -32 2 -16 3 -8 4 -4 5 -2 -1 0 6 2 4 10 9 7 0, 8 4 1 Rev. 0.2 8 11 16 12 64 32 13 14 15 AN278 Figure 5. Example DPCM Code Quantization To calculate the new predicted value for the second ADC sample, the DPCM code is reverted back to a difference value, which, for a DPCM code of 12, is a difference of 8. This difference is added back to the predicted value of 0x80 (128), which yields a new predicted value of 0x88 (136) for the next ADC sample. This is the same process that occurs during DPCM decompression, so that the same error is introduced on both sides. By using the decompression output (0x88) rather than the real sample (0x89), the compression and decompression schemes can include the same "DPCM" error and lessen the relative error. The predicted values in the compression scheme should match the predicted values and output of the decompression scheme exactly. DPCM code 1 2 3 4 5 6 7 0, 8 9 10 11 12 13 14 15 Difference -64 -32 -16 -8 -4 -2 -1 0 1 2 4 8 16 32 64 Figure 6. Example DPCM Decoding Scheme If the second ADC value is 0x87 (135), this is compared to the predicted value of 0x88 (136), which yields a difference of –1 and a DPCM code of 7, which is sent to memory. The new predicted value is the current predicted value 0x88 (136) added to the decoded DPCM value (–1), or 0x87 (135). The predicted value of 0x87 (135) will then be compared to the next ADC sample, and so on. With the compression quantization, any difference between 4 and 8 will yield a DPCM code of 11, a difference less than –64 will result in a DPCM code of 1, and a difference of 0 will result in a DPCM code of either 0 or 8. In the case of this DPCM algorithm implementation, the DPCM values are "snapped down" to the smaller difference value in the range bin for positive numbers and "snapped up" to the larger difference value in the range bin for negative numbers. For example, a difference of 15 uses the same code as a difference of 8. This is done to eliminate the chance of a DAC rollover upon play back. The following table and graph continue the DPCM algorithm on a set of example ADC samples. Notice how the DPCM algorithm follows the real ADC samples and is oftentimes close to the ADC sample. The more bits the DPCM algorithm uses relative to the number of ADC bits, the more accurate the results will be, as it allows for more differences to be represented by a DPCM code. Table 1. Example DPCM Algorithm Results Algorithm Example Iteration Real Samples Difference (Sample DPCM - Predicted_Value) Code DPCM Compression Predicted Decompression Decoded Values (Predicted_Value Output Difference + Decoded_Difference) 0 — — — — 128 128 1 137 9 12 8 136 136 2 135 –1 7 –1 135 135 3 138 3 10 2 137 137 4 140 3 10 2 139 139 5 132 –7 5 –4 135 135 6 120 –15 4 –8 127 127 7 100 –27 3 –16 111 111 8 107 –4 5 –4 107 107 9 111 4 11 4 111 111 Rev. 0.2 5 AN278 Table 1. Example DPCM Algorithm Results 6 10 114 3 10 2 113 113 11 113 0 0 0 113 113 12 110 –3 6 –2 111 111 13 112 1 9 1 112 112 Rev. 0.2 AN278 Figure 7. Example Input Waveform and DPCM Output During the DPCM decompression process, the first DPCM code of 12 is retrieved from memory, decoded to a difference of 8, and added to the initial predicted value of 0x80 to yield the output sample of 0x88 (136). The second DPCM code of 7 is then retrieved from memory, decoded to a difference of –1, and added to the predicted value of 0x88 (136) to yield the second output sample of 0x87, and so on. The Voice Recorder DPCM algorithm behaves the same as the above example, except the DPCM codes are 6 bits for a 12-bit ADC. Rev. 0.2 7 AN278 4.4. SPI Interface The SST Flash memory used in the voice recorder has various commands that consist of a specified op-code and operands. The datasheet for the Flash memory describes each command and the timing involved in great detail. The pertinent commands for this application are the byte-program writes, status register read and enable, read, and chip erase. Upon power-up, the Flash must have the block protection bits in the status register cleared in order to write to the memory. Additionally, every write and erase operation must be preceded by a write enable command. Furthermore, the timing for the /NSS line is critical, as the Flash will abort any command it has not fully received before /NSS is disabled. The software controlling the SPI uses both the TXBMT and SPIF flags to verify that all necessary transactions between the Flash and the 'F411 occur before disabling /NSS. 4.5. Suspend Mode The low power 'F411 includes a special Suspend mode feature that turns off the internal oscillator until a waking event occurs. The 'F411 controls a transistor that can shut off the power output from the external voltage regulator to conserve power in all other devices, as well. The lone red LED (PWR_LED) indicates whether the 'F411 has shut off power to all other devices. For the voice recorder, a port match event terminates the Suspend mode. This event can occur when the user presses one of the switches and the Port I/O, after a check with the port mask registers, mismatch the value in the port match registers. Once the 'F411 awakens, the peripherals and external SST Flash must be reinitialized after the memory's initial hold time of 10 µs. Normal voice recorder operation can then resume. 4.6. Detailed Descriptions Figure 8 depicts the main software routine. The voice recorder completes all system initializations and begins polling on the buttons. If a specified time period passes without any interaction, the system automatically switches to Suspend mode. When the user presses a switch, the recorder acknowledges the interaction by brightening an LED and the associated action takes place. When the action completes, the LED dims. 8 Rev. 0.2 AN278 System Initializations Memory already contains data? No Yes Update recording ending address Check switches Switch pressed? No No Yes Brighten LED 5 seconds passed? Execute switchspecific action Yes Enter SUSPEND mode Dim LED Turn off switch timer Figure 8. Initialization and Switch Polling Routine If the user presses the Record/Play push-button, the 'F411 first determines if the recorder should start recording (button held down) or playing (button pressed and released). Once recording, the 'F411 checks if the memory was erased and resets the beginning address of the new recording appropriately. The recorder then checks if the recording must end because the memory is full or because the user released the Record/Play push-button. If neither of these two conditions occur, then the recorder reads the sample from the ADC, executes the DPCM algorithm, and packs the 6-bit code into the byte to be sent to storage. If that byte now contains 8 bits of data, it's sent to memory; otherwise, the routine stores the byte until the next sampling time. This ISR is executed every 125 µs, or at a frequency of 8 kHz. Rev. 0.2 9 AN278 Was memory erased? No Yes Byte full? Restart at beginning of memory Yes No Yes Send to memory Memory full? No Yes Increment address Switch released? Decode sample No Encode ADC sample Update predictor Pack code into a databyte END Figure 9. Record Function and Sampling Interrupt Service Routine When the user plays the recording, as shown in Figure 10, the ISR first checks whether or not the playback should end because the user pressed the push-button again or the ISR reached the last memory address of the recording. If these checks prove to be false, the recorder unpacks a 6-bit code from the byte from memory, decodes it using the DPCM algorithm, updates the DAC output, and checks if a new byte should be fetched from memory. This ISR is also run every 125 µs. 10 Rev. 0.2 AN278 Yes Switch pressed again? Read packed code from memory No Yes End of recording reached? Increment Address No Data byte empty? Yes Unpack DPCM code from data byte No Decode sample Update output END Figure 10. Playback Interrupt Service Routine The other functions available to the recorder include recording erase and volume control. The erase uses the chiperase command available to the SST and is controlled in the switch-polling function, as it is a lower priority than record and playback. The volume control is accomplished through a potentiometer that changes the voltage divider at the IDAC output. Suspend mode, the final feature of the recorder, occurs after a set period of idle time. As shown in Figure 11, the recorder first turns off the timer controlling the push-button polling, for the functions associated with the switches should not be active until the 'F411 completes all re-initializations after exiting Suspend mode. The recorder also turns off all the peripherals and the power to the op-amps, external Flash memory, and speaker driver to conserve power. The C8051F411's internal oscillator halts while in Suspend mode until a waking event occurs. In this application, the relevant waking event is the port match, where the device will return to normal operation if one of the switches is pressed. After the port match event, the recorder reinitializes all peripherals, waits the power-up hold time specified for the SST Flash, initializes the SST Flash, and begins to poll the switches. Rev. 0.2 11 AN278 Turn off switch timer Turn off peripherals and external devices Wait for switch to be pressed (port match event) Turn on peripherals and external devices Wait until external Flash is ready Initialize external Flash Check switches Turn on switch timer Figure 11. Suspend Mode 12 Rev. 0.2 AN278 5. Usage Notes The voice recorder may be powered from either a 9 V battery or a 9 V DC Power Adapter. Protection diodes will stop any mishaps from occurring if both are in place at the same time. Each button activates and deactivates the corresponding function, and the LED will indicate whether the function is active. Certain functions can only be used during certain times; for example, the erase operation can only occur if no other action is taking place. Recording multiple times without erasing appends the new recording to the end of the first. The play function always begins at the very first recording. 6. Design Customization All of the polling, sampling periods, and DPCM quantization values are constants declared at the beginning of the software files. Any of the chosen values in this project may be changed, but take caution to observe that all requirements are met by each change. For example, changing the system clock divider may result in a sampling ISR that doesn't meet the 125 µs timing requirement, or changing the low-pass filter corner frequency may cause unwanted aliasing. 7. References Chipcon. "AN026: Wireless Audio using CC1010." Rev 1.0, 9/8/2004. Silicon Laboratories. "AN147: Wireless Digital Full-Duplex Voice Transceiver." Rev 1.1. SST. SST25VF040 Data Sheet. 6/04. Rev. 0.2 13 Figure 12. Voice Recorder Reference Design Schematic (Page 1 of 2) AN278 APPENDIX A—SCHEMATIC 14 Rev. 0.2 Figure 13. Voice Recorder Reference Design Schematic (Page 2 of 2) AN278 Rev. 0.2 15 AN278 APPENDIX B—BILL OF MATERIALS Table 2. Bill of Materials Qty Board Reference Value/Part Number Description Manufacturer 13 C2, C3, C5, C7, C11, C13, C16, C17, C20, C23, C25, C32, C33 0.1 µF 0805 any 3 C8, C14, C19 1.0 µF 0805 any 1 C29 1200 pF 0805 any 1 C24 1500 pF 0805 any 3 C27, C28, C30 2700 pF 0805 any 1 C31 4700 pF 0805 any 2 C9, C26 68000 pF 0805 any 1 C18 7 pF 0805 any 1 C34 8200 pF 0805 any 1 C10 ECA-1EHG331 330 µF Elect. Panasonic - ECG 3 C1, C4, C6 ECS-T0JY475R 4.7 µF Tant. Panasonic - ECG 1 C12 T491C156K010AS 15 µF Tant. Kemet 3 C15, C21, C22 T491B106K016AS 10 µF Tant. Kemet 1 R2 100 0805 any 2 R29, R30 100 k 0805 any 1 R19 12 k 0805 any 1 R15 133 k 0805 any 4 R16, R17, R26, R27 154 k 0805 any 1 R24 16.2 k 0805 any 4 R1, R4, R9, R10 1 k 0805 any 1 R7 2 0805 any 1 R21 2.7 k 0805 any 3 R5, R6, R8 200 0805 any 1 R18 24.3 k 0805 any 2 R3, R12 2 k 0805 any 1 R28 4.22 k 0805 any 2 R13, R23 4.99 k 0805 any 1 R14 5.1 M 0805 any 16 Rev. 0.2 AN278 Table 2. Bill of Materials (Continued) Qty Board Reference Value/Part Number Description Manufacturer 1 R20 6.34 k 0805 any 1 R25 7.68 k 0805 any 1 R22 9.76 k 0805 any 1 R11 3352T-1-501 500 potentiometer Bourns Inc. 1 BH1 1295 9 V battery holder Keystone Electronics 1 J1 103308-1 2 x 5 shrouded AMP/Tyco Electronics 1 J2 SJ-3543N audio jack CUI Inc. 1 J5 RAPC722 power jack Switchcraft Inc. 2 Z1, Z2 SD103CW-13 Schottky diode Diodes Inc. 1 D1 SML-LXT0805GW-TR green LED 0805 Lumex Opto/Components Inc. 2 D2, D4 SML-LXT0805IW-TR red LED 0805 Lumex Opto/Components Inc. 4 MH1, MH2, MH3, MH4 1902E stand-offs Keystone Electronics 1 MK1 EM6050P-443 Microphone Horn Industrial Co LTD 3 SW1, SW2, SW3 EVQ-PAD04M switches Panasonic - ECG 1 U1 C8051F411-GM QFN-28 Silicon Laboratories 1 U2 ZXMP3A13FTA transistor SOT23 Zetex Inc. 1 U3 SST25VF040B-50-4CQAF 512 kB Flash SST 1 U4 LM2936IMP-3.3 Vol. reg SOT223 National Semiconductor 1 U5 MC33204DR2 op-amp SOIC14 ON Semiconductor 1 U6 TPA4861D speaker driver SOIC8 Texas Instruments Rev. 0.2 17 AN278 APPENDIX C—LAYOUT Figure 14. Top Layout and Silkscreen Figure 15. Bottom Layout and Silkscreen 18 Rev. 0.2 AN278 APPENDIX D—SOFTWARE SOURCE CODE Startup Code (Modified STARTUP.A51) $NOMOD51 ;-----------------------------------------------------------------------------; This file is part of the C51 Compiler package ; Copyright (c) 1988-2001 Keil Elektronik GmbH and Keil Software, Inc. ;-----------------------------------------------------------------------------; STARTUP.A51: This code is executed after processor reset. ; ; To translate this file use A51 with the following invocation: ; ; A51 STARTUP.A51 ; ; To link the modified STARTUP.OBJ file to your application use the following ; BL51 invocation: ; ; BL51 <your object file list>, STARTUP.OBJ <controls> ; ;-----------------------------------------------------------------------------; ; User-defined Power-On Initialization of Memory ; ; With the following EQU statements the initialization of memory ; at processor reset can be defined: ; ; ; the absolute start-address of IDATA memory is always 0 IDATALEN EQU 80H ; the length of IDATA memory in bytes. ; XDATASTART EQU 0H ; the absolute start-address of XDATA memory XDATALEN EQU 0H ; the length of XDATA memory in bytes. ; PDATASTART EQU 0H ; the absolute start-address of PDATA memory PDATALEN EQU 0H ; the length of PDATA memory in bytes. ; ; Notes: The IDATA space overlaps physically the DATA and BIT areas of the ; 8051 CPU. At minimum the memory space occupied from the C51 ; run-time routines must be set to zero. ;-----------------------------------------------------------------------------; ; Reentrant Stack Initilization ; ; The following EQU statements define the stack pointer for reentrant ; functions and initialized it: ; ; Stack Space for reentrant functions in the SMALL model. IBPSTACK EQU 0 ; set to 1 if small reentrant is used. IBPSTACKTOP EQU 0FFH+1 ; set top of stack to highest location+1. ; ; Stack Space for reentrant functions in the LARGE model. XBPSTACK EQU 0 ; set to 1 if large reentrant is used. XBPSTACKTOP EQU 0FFFFH+1; set top of stack to highest location+1. ; ; Stack Space for reentrant functions in the COMPACT model. PBPSTACK EQU 0 ; set to 1 if compact reentrant is used. PBPSTACKTOP EQU 0FFFFH+1; set top of stack to highest location+1. ; ;-----------------------------------------------------------------------------; ; Page Definition for Using the Compact Model with 64 KByte xdata RAM ; Rev. 0.2 19 AN278 ; The following EQU statements define the xdata page used for pdata ; variables. The EQU PPAGE must conform with the PPAGE control used ; in the linker invocation. ; PPAGEENABLE EQU 0 ; set to 1 if pdata object are used. PPAGE EQU 0; define PPAGE number. ; ;-----------------------------------------------------------------------------; Standard SFR Symbols required in XBANKING.A51 ACC DATA 0E0H B DATA 0F0H SP DATA 81H DPL DATA 82H DPH DATA 83H PCA0MD DATA 0D9H NAME ?C_C51STARTUP ?STACK SEGMENT ?C_STARTUP CODE SEGMENT RSEG DS IDATA ?STACK 1 EXTRN CODE (?C_START) PUBLIC ?C_STARTUP ?C_STARTUP: LJMP CSEG AT0 STARTUP1 RSEG ?C_C51STARTUP STARTUP1: ANL IF IDATALEN <> 0 IDATALOOP: MOV ENDIF PCA0MD, #0BFH MOV CLR @R0,A DJNZ IF XDATALEN <> 0 MOV MOV IF (LOW (XDATALEN)) <> 0 MOV ELSE MOV ENDIF CLR XDATALOOP: MOVX @DPTR,A INC DJNZ DJNZ ENDIF IF PPAGEENABLE <> 0 ENDIF 20 MOV R0,#IDATALEN - 1 A R0,IDATALOOP DPTR,#XDATASTART R7,#LOW (XDATALEN) R6,#(HIGH (XDATALEN)) +1 R6,#HIGH (XDATALEN) A DPTR R7,XDATALOOP R6,XDATALOOP P2,#PPAGE Rev. 0.2 AN278 IF PDATALEN <> 0 PDATALOOP: MOVX ENDIF MOV MOV CLR @R0,A INC DJNZ R0,#PDATASTART R7,#LOW (PDATALEN) A R0 R7,PDATALOOP MOV ?C_IBP,#LOW IBPSTACKTOP MOV MOV ?C_XBP,#HIGH XBPSTACKTOP ?C_XBP+1,#LOW XBPSTACKTOP MOV ?C_PBP,#LOW PBPSTACKTOP IF IBPSTACK <> 0 EXTRN DATA (?C_IBP) ENDIF IF XBPSTACK <> 0 EXTRN DATA (?C_XBP) ENDIF IF PBPSTACK <> 0 EXTRN DATA (?C_PBP) ENDIF MOV SP,#?STACK-1 ; This code is required if you use L51_BANK.A51 with Banking Mode 4 ; EXTRN CODE (?B_SWITCH0) ; CALL ?B_SWITCH0 ; init bank mechanism to code bank 0 LJMP ?C_START END Rev. 0.2 21 AN278 Main Voice Recorder Program //----------------------------------------------------------------------------// F411_VR.c //----------------------------------------------------------------------------// Copyright 2006 Silicon Laboratories, Inc. // http://www.silabs.com // // Program Description: // // This program uses the DPCM functions to encode voice samples and saves them // to flash memory. This program also interfaces with a speaker or headphones // to play back the recorded voice. // // How To Use: See Readme.txt // // FID: 41X000005 // Target: C8051F411 // Tool chain: Keil C51 7.50 / Keil EVAL C51 // Silicon Laboratories IDE version 2.6 // Project Name: F411_VR // // Release 1.4 // -All changes by TP // -06 Feb 2009 // -Added a delay when restarting and waking up to allow the SST Flash // VDD to ramp up fully before initializing the Flash. // -Lengthened the delay before re-enabling the VDD Monitor // // Release 1.3 // -All changes by TP // -02 Feb 2006 // -minor changes in comments // // Release 1.2 // -All changes by TP // -21 Nov 2005 // -Revised for a 2-button version of the board with // volume wheel. // // Release 1.1 // -All changes by TP // -16 Aug 2004 // -project version updated, no changes to this file // // Release 1.0 // -Initial Revision (TP) // -15 AUG 2004 // //----------------------------------------------------------------------------// Includes //----------------------------------------------------------------------------#include <c8051f410.h> // SFR declarations #include "F411_VR_DPCM.h" // contains DPCM functions #include "F411_VR_SSTFlash.h" // contains functions to write to the // serial SST external 512 kb Flash #include "F411_VR_LED.h" // contains functions to control the // intensity of the LEDs //----------------------------------------------------------------------------// 16-bit SFR Definitions for 'F411 //----------------------------------------------------------------------------- 22 Rev. 0.2 AN278 // SFR16 Defintions (Timers, ADC, and DAC) sfr16 TMR2RL = 0xCA; // Timer 2 Reload address sfr16 TMR2 = 0xCC; // Timer 2 Counter address sfr16 TMR3RL = 0x92; // Timer 3 Reload address sfr16 TMR3 = 0x94; // Timer 3 Counter address sfr16 ADC0DAT = 0xBD; // ADC 16-bit address sfr16 IDA0DAT = 0x96; // IDAC 16-bit address //----------------------------------------------------------------------------// Global CONSTANTS //----------------------------------------------------------------------------// General Constants #define SYSCLK 6125000 // T1 runs at SYSCLK / 48 #define POLLING 1992 #define PRCHANGE 20000 // system clock in Hz (24.5 MHz / 4) // poll the switches at 64 Hz // wait approx. 150ms after a button is // pressed to "debounce" the switches #define SAMP_FREQ 764 // use 8kHz sampling for both the ADC // and DAC #define MAX_MEM_ADDR 0x0007FFFF #define NEAR_END_ADDR 0x00065F55 // // // // // // #define mid_range 2048 // middle value of 12-bit ADC and DAC // System States #define IDLE 0x00 #define RECORDING 0x01 #define PLAYING 0x02 #define END_MEM 0x04 #define ERASED 0x08 // // // // // // 512K = 2^19 address bits for every 4 samples, 3 bytes are written to memory, so 10.67 kHz ((8 kHz * 4)/3) writing time = 106666 addresses every 10 seconds, so give about 10 seconds of warning indicates indicates indicates flag used reached flag used no current the device the device if the end action is recording is playing of memory is if memory is erased // Port Pin Definitions sbit REC_PLAY = P1^7; sbit ERASE = P1^6; sbit TRANS = P1^3; sbit LED0 = P2^1; sbit LED1 = P2^0; sbit sbit sbit sbit SCK = P0^4; MISO = P0^5; MOSI = P0^6; NSS = P0^7; //----------------------------------------------------------------------------// Global VARIABLES //----------------------------------------------------------------------------unsigned char system_state = IDLE; // // // // // // // start in idle mode bit 3 of the system_state indicates if the end of memory has been reached bit 4 of the system_state indicates if the memory has been erased since the last action (1 = erased) Rev. 0.2 23 AN278 // Ending address of the recording in memory unsigned long rec_end_addr = 0x00000000; // flags to communicate between the T1 ISR and the ADC/DAC ISRs for various // termination events bit ADC_STOP_FLAG = 0; bit MEM_END_NEAR_FLAG = 0; bit MEM_END_FLAG = 0; bit REC_END_FLAG = 0; bit DAC_STOP_FLAG = 0; bit ENTER_SUSPEND = 0; //----------------------------------------------------------------------------// Function PROTOTYPES //----------------------------------------------------------------------------// System and peripheral initialization functions void System_Init (void); void VDDMon_Init (void); void Port_Init (void); void ADC0_Init (void); void DAC0_Init (void); void PCA_Init (void); void SPI0_Init (void); void RTC_Init (void); void Timer0_Init (int period); void Timer1_Init (int period); void Timer2_Init (int period); void Timer3_Init (int period); void Recording_Search (void); // Interrupt service routines void Timer0_ISR (void); void Timer1_ISR (void); void ADC0_ISR (void); void Timer3_ISR (void); // // // // LED updates Switch polling Recording Playback //----------------------------------------------------------------------------// MAIN Routine //----------------------------------------------------------------------------void main (void) { unsigned char i; // Watchdog timer disabled in VR_STARTUP.A51 System_Init (); Port_Init (); ADC0_Init (); DAC0_Init (); PCA_Init (); SPI0_Init(); 24 // // // // // // // Initialize Initialize Initialize Initialize Initialize in modules Initialize system clock crossbar and GPIO ADC0 (microphone) DAC0 (speaker) the PCA for 8-bit PWM 0, 1, and 2 the interface to the flash RTC_Init (); // Stop the RTC from causing a wake-up // from suspend Timer0_Init (LED_PWM); // Initialize timer 0 to provide a // 76 Hz interrupt rate for the LEDs Timer1_Init (POLLING); // Initialize timer 1 to provide a // 64 Hz interrupt rate for the switches Rev. 0.2 AN278 Timer2_Init (SAMP_FREQ); // Initialize the timer to provide an // 8KHz interrupt rate for sampling Timer3_Init (SAMP_FREQ); // Initialize the timer to provide an // 8KHz interrupt rate for sampling EA = 1; // enable global interrupts // wait for the VDD of the Flash to ramp up + 10us until the Flash // is ready to receive writes and reads for (i = 0; i < 255; i++); SSTFlash_Init (); // disable the write protection in the // external SST flash memory Recording_Search (); // search for a recording already // present in memory TR1 = 1; // start polling the switches // loop forever while (1) { if (ENTER_SUSPEND == 1) // check if no interaction has occurred { // for some time // disable everything to save the most power and set everything to a // dormant state ENTER_SUSPEND = 0; TR1 = 0; // stop polling the switches EA = 0; // disable all interrupts XBR1 = 0x40; XBR0 = 0x00; // disable the PCA // disable the SPI SCK = 0; MISO = 0; MOSI = 0; // drive the SPI pins low so the // external Flash won't attempt to draw // current while unpowered IDA0CN &= ~0x80; REF0CN &= ~0x01; // disable DAC0 // disable VREF LED0 = 1; LED1 = 1; // turn the LEDs off TRANS = 1; // turn off the external circuitry RSTSRC = 0x00; VDM0CN &= ~0x80; // disable missing clock detector // disable the VDD Monitor OSCICN |= 0x20; // enter suspend mode and wait // until a port match event occurs // re-enable and reinitialize the system VDDMon_Init (); TRANS = 0; // turn on the external circuitry REF0CN |= 0x01; IDA0CN |= 0x80; // re-enable VREF // re-enable DAC0 Rev. 0.2 25 AN278 XBR0 = 0x02; XBR1 = 0x42; // re-enable SPI // re-enable PCA0_0 and PCA0_1 // wait for the VDD of the Flash to ramp up + 10us until the Flash // is ready to receive writes and reads for (i = 0; i < 255; i++); SSTFlash_Init (); // re-initialize the SST flash EA = 1; // enable global interrupts // wait until the button that woke the system is released while ((REC_PLAY == 0) || (ERASE == 0)); } } } TR1 = 1; // begin polling the buttons again /////////////////////////// INITIALIZATION ROUTINES /////////////////////////// //----------------------------------------------------------------------------// System_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // This routine initializes the system clock to use the internal 24.5MHz / 4 // oscillator as its clock source and enables the missing clock detector reset. // Additionally, this routine sets up VREF, the internal regulator, and the // VDD monitor. // void System_Init (void) { OSCICN = 0x85; // configure internal oscillator RSTSRC = 0x04; // enable missing clock detector } REF0CN = 0x01; // set up and enable VREF pin REG0CN = 0x10; // set up and enable 2.5V VDD from the // internal regulator VDDMon_Init (); // initialize the VDD Monitor //----------------------------------------------------------------------------// VDDMon_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // This routine initializes the VDD Monitor and enables it as a reset source. // void VDDMon_Init (void) { unsigned char i; VDM0CN = 0x80; for (i = 0; i < 255; i++); RSTSRC = 0x06; } 26 // // // // enable the VDD monitor wait for the monitor to stabilize enable missing clock detector and VDD monitor as reset sources Rev. 0.2 AN278 //----------------------------------------------------------------------------// PORT_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // P0.0 = DAC0 (analog, skip) // P0.1-3 = unused (skip) // P0.4-7 = SPI interface (digital, do not skip) // P1.0-1 = unused (skip) // P1.2 = VREF (analog, skip) // P1.3 = analog power-on transistor (digital, skip) // P1.4 = unused (skip) // P1.5 = ADC0 (analog, skip) // P1.6-7 = REC_PLAY and ERASE switches (digital, skip) // P2.0-1 = LED PCA outputs (digital, do not skip) // void Port_Init (void) { P0MDIN = 0xFE; // make switch and SPI pins digital P0MDOUT = 0xD0; // make SPI pins push-pull P1MDIN = 0xC8; // make trans and switches digital P1MDOUT = 0x08; // make trans pin push-pull P2MDIN = 0x03; // make PCA pins digital P2MDOUT = 0x03; // make PCA pins push-pull P0SKIP = 0x0F; // skip pins not belonging to SPI P1SKIP = 0xFF; // skip all P1 pins P2SKIP = 0xFC; // skip pins not belonging to LED PCA } XBR0 = 0x02; XBR1 = 0x42; // enable SPI // enable PCA0_0 and PCA0_1 TRANS = 0; // turn on the power to all analog // components P0MAT = 0x00; P1MAT = 0xC0; P0MASK = 0x00; P1MASK = 0xC0; EIE2 = 0x00; // // // // // // the buttons will go low when pressed, causing the port match event mask off all P0 and P1 pins except the switches disable the port match interrupt (not required to wake up the core) //----------------------------------------------------------------------------// ADC0_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // Configure ADC0 to update with a Timer 2 overflow using P1.5 as its positive // input in post-tracking mode, enable burst mode, and use a repeat factor of // 16. // void ADC0_Init (void) { ADC0CN = 0x43; // ADC in low-power burst mode, use T2 // overflow, right justify ADC0MX = 0x0D; // use P1.5 as the positive reference // set the ADC conversion about 5 MHz and use a repeat factor // of 16 ADC0CF = (4 << 3) | (3 << 1); Rev. 0.2 27 AN278 ADC0TK = 0xF4; EIE1 |= 0x08; EIP1 |= 0x08; } // // // // // use post-tracking mode enable the ADC conversion complete interrupt set the ADC convertion complete interrupt to high priority //----------------------------------------------------------------------------// DAC0_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // Configure DAC0 to be right justified, update with a Timer 3 overflow, and // use a full-scale 2 mA output current. // void DAC0_Init (void) { IDA0CN = 0x70; // set the IDAC to update on a write // to IDA0DAT (initially only) IDA0CN |= 0x00; // set the IDAC to use a 0.25 mA current. IDA0CN |= 0x04; // set the IDAC to be right-justified IDA0CN |= 0x80; // enable the IDAC } IDA0L = 0x00; IDA0H = 0x08; // initialize the IDAC to be mid-scale IDA0CN &= ~0x70; IDA0CN |= 0x33; // set the IDAC to update on T3 overflow, // and use a 2 mA full-scale current //----------------------------------------------------------------------------// PCA_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // Configure PCA0 modules 0 and 1 to 8-bit PWM mode using the system clock. // void PCA_Init (void) { PCA0MD = 0x88; // set PCA to use system clock, disable // idle mode // PCA0 (for LED1) PCA0CPM0 = 0x42; PCA0CPH0 = 0x00; // set PCA0 for 8-bit PWM mode // set LED to off originally // PCA1 (for LED0) PCA0CPM1 = 0x42; PCA0CPH1 = 0x00; // set PCA1 for 8-bit PWM mode // set LED to off originally // add another PCA module for another LED here, if desired } PCA0CN = 0x40; // turn on the PCA timer/counter //----------------------------------------------------------------------------// SPI0_Init //----------------------------------------------------------------------------// // Return Value : None 28 Rev. 0.2 AN278 // Parameters : None // // Configure the SPI to run in 4-wire master mode at SYSCLK / 4 (1.53 MHz) // using clock phase 0 and clock polarity 0 to interface with the SST Flash // memory. // void SPI0_Init (void) { SPI0CFG = 0x40; // set the master mode, polarity and // phase // set the SPI frequency to SYSCLK / 2*(1+1) = SYSCLK / 4 SPI0CKR = 0x01; SPI0CN = 0x0C; } SPIEN = 1; // clear flags, turn off NSS // set the 4-wire mode // enable the SPI //----------------------------------------------------------------------------// Timer0_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : // 1) int period - number of timer counts to generate the desired period // range is postive range of integer: 0 to 32767 // // Configure Timer0 to 16-bit mode. Timer0 is used to control the load // time of the PCA PCA0CPHn registers, which changes the PWM intensity of the // LEDs. // // The input parameter can be calculated as follows: // (Oscillator (Hz) / 4) / Desired_Freq (Hz) = Timer Ticks // void Timer0_Init (int period) { TMOD |= 0x01; // set Timer 0 to mode 1 (16 bit) CKCON |= 0x04; // use the system clock ET0 = 1; // enable Timer 0 interrupts PT0 = 1; // set Timer 0 interrupts to high // priority (has to interrupt T1) } TL0 = (-period) & 0x00FF; TH0 = ((-period) & 0xFF00) >> 8; // set the desired period TR0 = 0; // keep Timer 0 off (LED // functions will turn it on) //----------------------------------------------------------------------------// Timer1_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : // 1) int period - number of timer counts to generate the desired period // range is postive range of integer: 0 to 32767 // // Configure Timer1 to 16-bit mode. Timer1 controls the switch polling. // // To calculate: // (Oscillator (Hz) / 4) / 48 / Desired_Freq (Hz) = Timer Ticks // // NOTE - the extra 48 in this equation is present because of the settings Rev. 0.2 29 AN278 // in CKCON. // void Timer1_Init (int period) { TMOD |= 0x10; CKCON |= 0x02; ET1 = 1; } // set Timer 1 to mode 1 (16 bit) // use the system clock / 48 // enable Timer 1 interrupts TL1 = (-period) & 0x00FF; TH1 = ((-period) & 0xFF00) >> 8; // set the desired period TR1 = 0; // keep Timer 1 off until needed //----------------------------------------------------------------------------// Timer2_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : // 1) int period - number of timer counts to generate the desired period // range is postive range of integer: 0 to 32767 // // Configure Timer2 to 16-bit auto reload mode. Timer2 controls the ADC0 // start-of-conversion rate. // // To calculate: // (Oscillator (Hz) / 4) / Desired_Freq (Hz) = Timer Ticks // void Timer2_Init (int period) { CKCON |= 0x10; // use the system clock TMR2CN = 0x00; // 16-bit auto-reload mode ET2 = 0; // disable T2 interrupts (use ADC // conversion complete interrupt) } TMR2RL = -period; // set the desired period TMR2 = -period; // initialize the timer TR2 = 0; // keep Timer 2 off until the RECORD // function is used //----------------------------------------------------------------------------// Timer3_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : // 1) int period - number of timer counts to generate the desired period // range is postive range of integer: 0 to 32767 // // Configure Timer3 to 16-bit auto reload mode. Timer3 controls the DAC output // rate. // // To calculate: // (Oscillator (Hz) / 4) / Desired_Freq (Hz) = Timer Ticks // void Timer3_Init (int period) { CKCON |= 0x40; // use the system clock TMR3CN = 0x00; // 16-bit auto-reload mode EIE1 |= 0x80; // enable Timer 3 interrupts 30 Rev. 0.2 AN278 } EIP1 |= 0x80; // set Timer 3 interrupts to high // priority TMR3RL = -period; // set the desired period TMR3 = -period; // initialize the timer TMR3CN = 0x00; // keep Timer 3 off until the PLAY // function is used //----------------------------------------------------------------------------// RTC_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // Enable the RTC so it doesn't cause a wake-up from suspend mode. // void RTC_Init (void) { RTC0KEY = 0xA5; // unlock the RTC interface RTC0KEY = 0xF1; RTC0ADR = 0x06; // point to RTC0CN RTC0DAT = 0x80; // enable the RTC } //----------------------------------------------------------------------------// Recording_Search //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // Search for a recording already residing in memory on power-up and set the // rec_end_addr accordingly. // void Recording_Search(void) { unsigned long address = 0x00000000; bit end_flag = 0; // indicate to the user that the microcontroller is not ready to record // or playback LED_DCH = &LED0_DC; Brighten_LED (); LED_DCH = &LED1_DC; Brighten_LED (); // search through the SST flash until a series of 0xFF is found, indicating // cleared memory while (end_flag != 1) { if (Read_MEM_Init (address) == 0xFF) { // double-check that the 0xFF found is not just a data byte of 0xFF if (Read_MEM_Init (address+10) == 0xFF) { if (Read_MEM_Init (address+40) == 0xFF) { end_flag = 1; } } Rev. 0.2 31 AN278 } address++; } if (address == MAX_MEM_ADDR) { end_flag = 1; } rec_end_addr = address-1; } // turn LED_DCH Dim_LED LED_DCH Dim_LED // set the recording ending address off the LEDs so the user knows the recording search has ended = &LED0_DC; (); = &LED1_DC; (); ////////////////////////// INTERRUPT SERVICE ROUTINES ///////////////////////// //----------------------------------------------------------------------------// Timer0_ISR //----------------------------------------------------------------------------// // Handle the 76Hz (13ms) Timer 0 interrupt. // // Timer 0 controls the rate at which the microcontroller changes the duty // cycle of the PCA controlling the LEDs // // The LEDs are updated periodically, even if the LED PWM hasn't changed. // By using the pointer (which is set before calling the LED functions) and // updating all LEDs in the ISR every time, the same functions can be used for // any number of LEDs. To add an LED, simply set-up another PCA channel, // point to that LED before calling the LED functions, and update the LED in // the ISR. // void Timer0_ISR (void) interrupt 1 using 1 { *LED_DCH += ADJ; // calculate the new duty cycle based // on the values set by the LED // functions PCA0CPH1 = LED0_DC; // load all LEDs with the possibly PCA0CPH0 = LED1_DC; // updated value // add another LED update here, if desired } TL0 = (-LED_PWM) & 0x00FF; TH0 = ((-LED_PWM) & 0xFF00) >> 8; // wait the time specified by the // calling LED function LED_PWM += LED_PWM_CHANGE; // change the interrupt rate, if // necessary //----------------------------------------------------------------------------// Timer1_ISR //----------------------------------------------------------------------------// // Handle the 64 Hz (15.63 ms) Timer 1 interrupt. // // Timer 1 controls the rate at which the microcontroller checks the switches // for activity while in full power mode. // // for RECORD - press and hold REC_PLAY button, release stops recording 32 Rev. 0.2 AN278 // for PLAYBACK - press and release REC_PLAY button, press and release again // to stop // void Timer1_ISR (void) interrupt 3 using 0 { // interrupt again in 15.63 ms, unless a switch is pressed unsigned short reload_value = POLLING; static unsigned char record_counter = 0; static unsigned short suspend_counter = 0; bit switch_pressed_flag = 0; // REC_PLAY button pressed if (REC_PLAY == 0) { switch_pressed_flag = 1; // record the user interaction // check if the recording time ran out, and stop any interaction // from the switch until the switch is released and pressed again if ((system_state & END_MEM) != END_MEM) { // the REC_PLAY button must be pressed and held for a period of time // in order to start the RECORD function record_counter++; // // // // if { check if the REC_PLAY button was held down long enough to begin recording (7 x 150 ms = 1.5 seconds) ignore the ERASED and END_MEM state bits, check if the system is idle and can start recording ((record_counter > 7) && ((system_state & 0x03) == IDLE)) TR2 = 1; // turn on the RECORD timer system_state |= RECORDING; // start recording LED_DCH = &LED0_DC; } Brighten_LED(); // point to the record LED's duty cycle // address // ramp on the record LED record_counter = 0; // reset the counter reload_value = PRCHANGE*2; // give a longer time period to check // the button (effectively debouncing) // check if the recording time is running out (button must be held // to continue recording) if (TR2 == 1) { if (MEM_END_NEAR_FLAG == 1) { LED_DCH = &LED0_DC; Flutter_LED (); // indicate to the user that time is // almost out } } else { // check if end of the memory has been reached if (MEM_END_FLAG == 1) { // stop recording system_state = IDLE | END_MEM; // indicate that the end of // memory was reached Rev. 0.2 33 AN278 MEM_END_FLAG = 0; LED_DCH = &LED0_DC; } } } Dim_LED (); // point to the record LED's duty cycle // address // dim off the record LED } else { // check if the switch was pressed, but not long enough to start // recording if (record_counter > 0) { switch_pressed_flag = 1; // record the user interaction // the system is currently playing - stop playing // ignore the ERASED and END_MEM state bits if ((system_state & 0x03) == PLAYING) { system_state &= ~PLAYING; // clear the PLAYING state bit DAC_STOP_FLAG = 1; IDA0DAT = 0x0800; LED_DCH = &LED1_DC; // point to the play LED's duty cycle // address // dim off the play LED Dim_LED (); } else { // the system is idle - start playing // ignore the ERASED and END_MEM state bits if ((system_state & 0x03) == IDLE) { system_state |= PLAYING; TMR3CN = 0x04; // start the timer controlling the DAC REC_END_FLAG = 0; // reset the "end of recording" flag DAC_STOP_FLAG = 0; LED_DCH = &LED1_DC; } } Brighten_LED (); record_counter = 0; // point to the play LED's duty cycle // address // ramp on the play LED // switch-press registered, reset } // the REC_PLAY switch was not pressed else { // clear the END_MEM recording flag after the ADC ISR has turned off // the ADC if ((system_state & END_MEM) == END_MEM) { system_state &= ~RECORDING; } // the system is currently recording - stop recording if (system_state == RECORDING) { system_state &= ~RECORDING; ADC_STOP_FLAG = 1; // notify the ADC to stop recording 34 Rev. 0.2 AN278 } MEM_END_NEAR_FLAG = 0; MEM_END_FLAG = 0; // clear all flags LED_DCH = &LED0_DC; // point to the record LED's duty cycle // address // dim off the record LED Dim_LED (); // check if the playback has reached the end of the recording if (REC_END_FLAG == 1) { // stop playing system_state &= ~PLAYING; REC_END_FLAG = 0; LED_DCH = &LED1_DC; } } } Dim_LED (); // point to the play LED's duty cycle // address // dim off the play LED // ERASE button pressed if (ERASE == 0) { // do nothing if the device is currently recording or playing // ignore the ERASED and END_MEM bits if ((system_state & 0x03) == IDLE) { // Indicate to the user that the microcontroller is busy LED_DCH = &LED1_DC; Brighten_LED (); LED_DCH = &LED0_DC; Brighten_LED (); rec_end_addr = 0x00000000; system_state |= ERASED; Erase_MEM (); } } LED_DCH Dim_LED LED_DCH Dim_LED // reset the counter // set the erase bit // erase the external SST Flash = &LED1_DC; (); = &LED0_DC; (); switch_pressed_flag = 1; // record the user interaction if (switch_pressed_flag == 0) { // check if the recorder is sitting and idle // ignore the ERASED and END_MEM bits if ((system_state & 0x03) == IDLE) { suspend_counter++; } // if no interaction occurs in 5 seconds, enter suspend mode if (suspend_counter == 320) { suspend_counter = 0; ENTER_SUSPEND = 1; } Rev. 0.2 35 AN278 } else { suspend_counter = 0; } } reload_value = PRCHANGE; // reset the SUSPEND mode counter // if the user is interacting with the // recorder // interrupt again in 150 ms // reload the timer for the next interrupt TL1 = (-reload_value) & 0x00FF; TH1 = ((-reload_value) & 0xFF00) >> 8; //----------------------------------------------------------------------------// ADC0_ISR //----------------------------------------------------------------------------// // Handle the 8kHz Timer 2 interrupt. // // Timer 2 controls the rate at which the ADC samples the input (RECORD). // void ADC0_ISR (void) interrupt 10 using 2 { // RECORD // DPCM variables static data unsigned short predicted_value = mid_range; static data unsigned char packed_code = 0x00; data unsigned char dpcm_code = 0x00; // indicates how the current dpcm_code should be packed to be sent to memory // sample 1 dpcm_code = A, sample 2 dpcm_code = B, sample 3 dpcm_code = C // sample 4 dpcm_code = D, sample 5 is the same as sample 1, etc // [A|A|A|A|A|A|B|B] = byte 1 // [B|B|B|B|C|C|C|C] = byte 2 // [C|C|D|D|D|D|D|D] = byte 3 static unsigned char state = 0; static short sample = 0x0000; AD0INT = 0; // clear the interrupt flag // check if the memory was erased if ((system_state & ERASED) == ERASED) { system_state &= ~ERASED; // clear the erased bit predicted_value = mid_range; // reset the dpcm predictor state = 0; // reset the packing state machine } // check for the end of memory if (rec_end_addr == MAX_MEM_ADDR) { TR2 = 0; // MEM_END_NEAR_FLAG = 0; MEM_END_FLAG = 1; // predicted_value = mid_range; // state = 0; // } else { // check if the REC_PLAY switch was 36 turn off T2 tell the T1 ISR to turn off the LED reset the dpcm predictor reset the state machine released and the recording should Rev. 0.2 AN278 // stop if (ADC_STOP_FLAG == 1) { TR2 = 0; ADC_STOP_FLAG = 0; // turn off T2 // reset the flag // do not reset the state or the predicted_value variables here // the playback ISR doesn't know when a recording starts or ends, // so it will also not reset the state and predicted_value } // take the sample, average it, compress it, and send it to memory else { // since 16 samples are automatically accumulated by the ADC, // average them by dividing by 16 (right shifting by 4) sample = (ADC0DAT >> 4) & 0x0FFF; // calculate the difference between the sample and the predictor // and compress the sample to a 6-bit DPCM code dpcm_code = DPCM_Encode ((sample - predicted_value)); // pack the DPCM code into the bytes sent to the Flash memory switch (state) { // state machine: 0 -> 1 -> 2 -> 3 // ^______________| case 0: // move the DPCM code into the 6 high bits // [A|A|A|A|A|A| | ] = byte 1 packed_code = (dpcm_code << 2) & 0xFC; state = 1; break; case 1: // move the DPCM code into the 2 low bits // of the previously packed byte // [-|-|-|-|-|-|B|B] = byte 1 packed_code |= (dpcm_code >> 4) & 0x03; Write_MEM (rec_end_addr, packed_code); rec_end_addr++; // move the rest of the DPCM code into the // 4 high bits of the next packed byte // [B|B|B|B| | | | ] = byte 2 packed_code = (dpcm_code << 4) & 0xF0; state = 2; break; case 2: // move the next DPCM code into the // 4 low bits of the previously packed byte // [-|-|-|-|C|C|C|C] = byte 2 packed_code |= (dpcm_code >> 2) & 0x0F; Rev. 0.2 37 AN278 Write_MEM (rec_end_addr, packed_code); rec_end_addr++; // move the rest of the DPCM code into the // 2 high bits of the next packed byte // [C|C| | | | | | ] = byte 3 packed_code = (dpcm_code << 6) & 0xC0; state = 3; break; case 3: // move the next DPCM code into the // 6 low bits of the previously packed byte // [-|-|D|D|D|D|D|D] = byte 3 packed_code |= dpcm_code & 0x3F; Write_MEM (rec_end_addr, packed_code); rec_end_addr++; state = 0; break; } default: state = 0; break; // indicate that the T1 ISR should flutter the LED, since the end of // memory is close if (rec_end_addr == NEAR_END_ADDR) { MEM_END_NEAR_FLAG = 1; } } } } // update the predictor for the next sample predicted_value += DPCM_Decode (dpcm_code); //----------------------------------------------------------------------------// Timer3_ISR //----------------------------------------------------------------------------// // Handle the 8kHz Timer 3 interrupt. // // Timer 3 controls the rate at which the DAC outputs decompressed samples // (PLAY). // void Timer3_ISR(void) interrupt 14 using 2 { // PLAY // next unwritten address static unsigned long current_play_addr = 0x00000000; // DPCM variables static unsigned short predicted_value = mid_range; static unsigned char packed_code = 0x00; 38 Rev. 0.2 AN278 unsigned char dpcm_code = 0x00; // indicates how the current dpcm_code should be unpacked when retrieved // from memory // sample 1 dpcm_code = A, sample 2 dpcm_code = B, sample 3 dpcm_code = C // sample 4 dpcm_code = D, sample 5 is the same as sample 1, etc // [A|A|A|A|A|A|B|B] = byte 1 // [B|B|B|B|C|C|C|C] = byte 2 // [C|C|D|D|D|D|D|D] = byte 3 static unsigned char state = 0; TMR3CN &= 0x7F; // clear the T3 interrupt flag // check if the PLAY switch was pressed and playing should stop if (DAC_STOP_FLAG == 1) { TMR3CN = 0x00; // turn off T3 DAC_STOP_FLAG = 0; // reset the flag current_play_addr = 0x00000000; // start at the beginning address predicted_value = mid_range; // reset the predictor state = 0; // reset the playback state machine } else { // check for the end of the recording if (current_play_addr >= rec_end_addr) { TMR3CN = 0x00; // turn off the timer REC_END_FLAG = 1; // tell the T1 ISR to turn off the LED current_play_addr = 0x00000000; predicted_value = mid_range; // reset the predictor state = 0; // reset the playback state machine } else { // unpack the DPCM code bytes retrieved from memory switch (state) { // state machine: 0 -> 1 -> 2 -> 3 // ^______________| case 0: packed_code = Read_MEM (current_play_addr); current_play_addr++; // take the DPCM code from the 6 high bits // [A|A|A|A|A|A| | ] = byte 1 dpcm_code = (packed_code >> 2) & 0x3F; state = 1; break; case 1: // take the next DPCM code from the 2 low bits // of the previously retrieved byte // [-|-|-|-|-|-|B|B] = byte 1 dpcm_code = (packed_code << 4) & 0x30; packed_code = Read_MEM (current_play_addr); current_play_addr++; // take the rest of the DPCM code from the // 4 high bits of the next retrieved byte Rev. 0.2 39 AN278 // [B|B|B|B| | | | ] = byte 2 dpcm_code |= (packed_code >> 4) & 0x0F; state = 2; break; case 2: // take the next DPCM code from the // 4 low bits of the previously retrieved byte // [-|-|-|-|C|C|C|C] = byte 2 dpcm_code = (packed_code << 2) & 0x3C; packed_code = Read_MEM (current_play_addr); current_play_addr++; // take the rest of the DPCM code from the // 2 high bits of the next retrieved byte // [C|C| | | | | | ] = byte 3 dpcm_code |= (packed_code >> 6) & 0x03; state = 3; break; case 3: // take the next DPCM code from the // 6 low bits of the previously retrieved byte // [-|-|D|D|D|D|D|D] = byte 3 dpcm_code = packed_code & 0x3F; state = 0; break; } default: state = 0; break; // calculate the new predicted value predicted_value += DPCM_Decode (dpcm_code); // output the new sample to the speaker IDA0DAT = predicted_value; // // // // if { } } } } overwrite the very last sample so the output is at the mid-range when stopped the discontinuity causes a small "clicking" sound when playback starts and stops (current_play_addr >= rec_end_addr) IDA0DAT = mid_range; //----------------------------------------------------------------------------// End Of File //----------------------------------------------------------------------------- 40 Rev. 0.2 AN278 External Flash Access Functions //----------------------------------------------------------------------------// F411_VR_SSTFlash.c //----------------------------------------------------------------------------// Copyright 2006 Silicon Laboratories, Inc. // http://www.silabs.com // // Program Description: // // This file contains the interfacing functions to the SST Flash, allowing the // user to Read memory, Write memory, and Erase memory. // // How To Use: See Readme.txt // // FID: 41X000010 // Target: C8051F411 // Tool chain: Keil C51 7.50 / Keil EVAL C51 // Silicon Laboratories IDE version 2.6 // Project Name: F411_VR // // Release 1.4 // -All changes by TP // -06 Feb 2009 // -No changes to this file // // Release 1.3 // -All changes by TP // -02 Feb 2006 // -added Read_MEM_Init (duplicate of Read_MEM) to avoid // the compiler warning (multiple calls to segment) // // Release 1.2 // -All changes by TP // -21 Nov 2005 // -project version updated, no changes to this file. // // Release 1.1 // -All changes by TP // -16 Aug 2004 // -added SPIF polling while sending the WREN command // // Release 1.0 // -Initial Revision (TP) // -15 AUG 2004 // //----------------------------------------------------------------------------// Includes //----------------------------------------------------------------------------#include <c8051f410.h> // SFR declarations //----------------------------------------------------------------------------// Global CONSTANTS //----------------------------------------------------------------------------// SST Instruction Opcodes (as shown in the datasheet) #define EWSR 0x50 // enable write status register #define WRSR 0x01 // write status register #define RDSR 0x05 // read status register #define WREN 0x06 // write enable #define BPROG 0x02 // byte program #define READ 0x03 // read #define CERASE 0x60 // chip erase #define READID 0x90 // chip ID Rev. 0.2 41 AN278 // Address definition typedef union ADDRESS { // access an address as a unsigned long ULong; // unsigned long variable or unsigned char UByte[4]; // 4 unsigned byte variables // [0] = A31-24, [1] = A23-16, [2] = A15-8, [3] = A7-0 } ADDRESS; //----------------------------------------------------------------------------// Function PROTOTYPES //----------------------------------------------------------------------------void SSTFlash_Init (void); void Write_MEM (unsigned long address, unsigned char data_byte); unsigned char Read_MEM (unsigned long address); void Erase_MEM (void); char ReadID_MEM (void); //----------------------------------------------------------------------------// SSTFlash_Init //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // Unprotect the memory so that all of memory may be written and read. // NOTE: The SPI must be initialized before this function is called. // void SSTFlash_Init (void) { NSSMD0 = 0; // enable the flash // send the enable write status register command SPI0DAT = EWSR; // load the XMIT register while (TXBMT != 1) // wait until EWSR command is moved into { // the XMIT buffer } SPIF = 0; while (SPIF != 1) // wait until the SPI finishes sending { // the EWSR command to the flash } SPIF = 0; NSSMD0 = 1; // allow the command to execute NSSMD0 = 0; // enable the flash // send the write status register command and clear the BP bits SPI0DAT = WRSR; // load the XMIT register while (TXBMT != 1) // wait until the XMIT register can { // accept more data } SPI0DAT = 0x00; // set the block protection bits to 0 while (TXBMT != 1) // wait until the data is moved into { // the XMIT buffer } SPIF = 0; while (SPIF != 1) // wait until the SPI finishes sending { // the data to the flash } SPIF = 0; } 42 NSSMD0 = 1; // allow the command to execute Rev. 0.2 AN278 //----------------------------------------------------------------------------// Write_MEM //----------------------------------------------------------------------------// // Return Value : None // Parameters : // 1) long address - address in the 512 kB external SST Flash // range is postive values up to 2^19: 0 to 524287, // or, 0 to 0x7FFFF // 2) char data_byte - the data to be written to memory // range is positive range of character: 0 to 255 // // Write one byte of data to a 24-bit address in the SST Flash Memory using // the SPI. // void Write_MEM (unsigned long address, unsigned char data_byte) { ADDRESS temp_addr; temp_addr.ULong = address; NSSMD0 = 0; // send the write enable command SPI0DAT = WREN; while (TXBMT != 1) { } SPIF = 0; while (SPIF != 1) { } SPIF = 0; // enable the flash // load the XMIT register // wait until the command is moved into // the XMIT buffer // wait until the command reaches the // flash NSSMD0 = 1; // allow the WREN to execute NSSMD0 = 0; // enable the flash // send the byte-program command SPI0DAT = BPROG; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[1]; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[2]; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[3]; while (TXBMT != 1) { } SPI0DAT = data_byte; while (TXBMT != 1) { } SPIF = 0; while (SPIF != 1) { } SPIF = 0; // load the XMIT register // wait until the command is moved into // the XMIT buffer // load the high byte of the address // wait until the addr is moved into // the XMIT buffer // load the middle byte of the address // wait until the addr is moved into // the XMIT buffer // load the low byte of the address // wait until the addr is moved into // the XMIT buffer // load the byte of data // wait until the data is moved into // the XMIT buffer // wait until the last byte of the // write instruction reaches the flash Rev. 0.2 43 AN278 } NSSMD0 = 1; // allow the WR instruction to execute //----------------------------------------------------------------------------// Read_MEM //----------------------------------------------------------------------------// // Return Value : // 1) char data_byte - the data byte read from memory // range is positive range of character: 0 to 255 // Parameters : // 1) long address - address in the 512 kB external SST Flash // range is postive values up to 2^19: 0 to 524287, // or, 0 to 0x7FFFF // // Read one byte of data from a 24-bit address in the SST Flash Memory using // the SPI. // unsigned char Read_MEM (unsigned long address) { ADDRESS temp_addr; temp_addr.ULong = address; NSSMD0 = 0; // send the read instruction SPI0DAT = READ; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[1]; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[2]; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[3]; while (TXBMT != 1) { } SPI0DAT = 0xFF; while (TXBMT != 1) { } SPIF = 0; while (SPIF != 1) { } SPIF = 0; NSSMD0 = 1; } // enable the flash // load the XMIT register // wait until the command is moved into // the XMIT buffer // load the high byte of the address // wait until the data is moved into // the XMIT buffer // load the middle byte of the address // wait until the data is moved into // the XMIT buffer // load the low byte of the address // wait until the data is moved into // the XMIT buffer // // // // load data wait into junk data in order to receive from the flash until the junk data is moved the XMIT buffer // wait until the read data is received // disable the flash return SPI0DAT; //----------------------------------------------------------------------------// Erase_MEM //----------------------------------------------------------------------------// // Return Value : None 44 Rev. 0.2 AN278 // Parameters : None // // Erase all data from the SST flash memory. // void Erase_MEM (void) { unsigned char mem_status = 0x01; NSSMD0 = 0; // send the write enable command SPI0DAT = WREN; while (TXBMT != 1) { } SPIF = 0; while (SPIF != 1) { } SPIF = 0; // enable the flash // load the XMIT register // wait until the command is moved into // the XMIT buffer // wait until the command reaches the // flash NSSMD0 = 1; // allow the WREN to execute NSSMD0 = 0; // enable the flash // send the chip erase instruction SPI0DAT = CERASE; while (TXBMT != 1) { } SPIF = 0; while (SPIF != 1) { } SPIF = 0; NSSMD0 = 1; // load the XMIT register // wait until the command is moved into // the XMIT buffer // wait until the command reaches the // flash // allow the erase to execute // poll on the busy bit in the flash until the erase operation is complete NSSMD0 = 0; // enable the flash SPI0DAT = RDSR; // send the read status register command while (TXBMT != 1) // wait until the SPI can accept more { // data } while (mem_status == 0x01) { SPI0DAT = 0xFF; // send junk in order to receive data while (TXBMT != 1) // wait until the junk data is moved { // into the XMIT buffer } SPIF = 0; while (SPIF != 1) // wait until the read data is received { } SPIF = 0; mem_status = SPI0DAT & 0x01; // check the BUSY bit } } NSSMD0 = 1; // disable the flash //----------------------------------------------------------------------------// Read_MEM_Init //----------------------------------------------------------------------------- Rev. 0.2 45 AN278 // // Return Value : // 1) char data_byte - the data byte read from memory // range is positive range of character: 0 to 255 // Parameters : // 1) long address - address in the 512 kB external SST Flash // range is postive values up to 2^19: 0 to 524287, // or, 0 to 0x7FFFF // // Read one byte of data from a 24-bit address in the SST Flash Memory using // the SPI. This function is called by Recording_Search in F411_VR.c // and is a duplicate of Read_MEM to avoid a warning by the compiler. // unsigned char Read_MEM_Init (unsigned long address) { ADDRESS temp_addr; temp_addr.ULong = address; NSSMD0 = 0; // send the read instruction SPI0DAT = READ; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[1]; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[2]; while (TXBMT != 1) { } SPI0DAT = temp_addr.UByte[3]; while (TXBMT != 1) { } SPI0DAT = 0xFF; while (TXBMT != 1) { } SPIF = 0; while (SPIF != 1) { } SPIF = 0; NSSMD0 = 1; } // enable the flash // load the XMIT register // wait until the command is moved into // the XMIT buffer // load the high byte of the address // wait until the data is moved into // the XMIT buffer // load the middle byte of the address // wait until the data is moved into // the XMIT buffer // load the low byte of the address // wait until the data is moved into // the XMIT buffer // // // // load data wait into junk data in order to receive from the flash until the junk data is moved the XMIT buffer // wait until the read data is received // disable the flash return SPI0DAT; //----------------------------------------------------------------------------// ReadID_MEM //----------------------------------------------------------------------------// // Return Value : // 1) char data_byte - the device ID read from memory at address 0x000001 // (this address is specified in the SST Flash datasheet) // range is positive range of character: 0 to 255 // Parameters : None // // Read the part ID from the flash memory (used for debugging). 46 Rev. 0.2 AN278 // char ReadID_MEM (void) { NSSMD0 = 0; } // enable the flash SPI0DAT = READID; while (TXBMT != 1) { } SPI0DAT = 0x00; while (TXBMT != 1) { } SPI0DAT = 0x00; while (TXBMT != 1) { } SPI0DAT = 0x01; while (TXBMT != 1) { } SPI0DAT = 0xA5; while (TXBMT != 1) { } SPIF = 0; while (SPIF != 1) { } SPIF = 0; // send the read ID instruction // wait until the SPI can accept more // data NSSMD0 = 1; // disable the flash // send the device ID address // wait until the SPI can accept more // data // send the device ID address // wait until the SPI can accept more // data // send the device ID address // wait until the SPI can accept more // data // send dummy data for shift register // wait until the SPI can accept more // data // wait until the read data is received return SPI0DAT; //----------------------------------------------------------------------------// End Of File //----------------------------------------------------------------------------- Rev. 0.2 47 AN278 DPCM (Differential Pulse Code Modulation) Functions //----------------------------------------------------------------------------// F411_VR_DPCM.c //----------------------------------------------------------------------------// Copyright 2006 Silicon Laboratories, Inc. // http://www.silabs.com // // Program Description: // // This file contains the DPCM encoding and decoding functions. // // NOTE: For another reference for DPCM, please see Chipcon's app note an026. // // NOTE: The calling function must have the same register context as the DCPM // functions, so it must either have the keyword "using 2" or all "using 2" // keywords for the DPCM functions need to be removed // // How To Use: See Readme.txt // // FID: 41X000006 // Target: C8051F411 // Tool chain: Keil C51 7.50 / Keil EVAL C51 // Silicon Laboratories IDE version 2.6 // Project Name: F411_VR // // Release 1.4 // -All changes by TP // -06 Feb 2009 // -No changes to this file // // Release 1.3 // -All changes by TP // -02 Feb 2006 // -project version updated, no changes to this file // // Release 1.2 // -All changes by TP // -21 Nov 2005 // -expanded the 4-bit codes to 6 bits for the 12-bit ADC // // Release 1.1 // -All changes by TP // -16 Aug 2004 // -project version updated, no changes to this file // // Release 1.0 // -Initial Revision (TP) // -15 AUG 2004 // //----------------------------------------------------------------------------// Includes //----------------------------------------------------------------------------#include <c8051f410.h> // SFR declarations //----------------------------------------------------------------------------// Global CONSTANTS //----------------------------------------------------------------------------// 12-bit quantization codes (6 bits, so 64 codes total = 31 positive, 31 // negative, and 2 zeroes) #define quant1 1 #define quant2 2 48 Rev. 0.2 AN278 #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define #define quant3 quant4 quant5 quant6 quant7 quant8 quant9 quant10 quant11 quant12 quant13 quant14 quant15 quant16 quant17 quant18 quant19 quant20 quant21 quant22 quant23 quant24 quant25 quant26 quant27 quant28 quant29 quant30 quant31 4 7 11 16 22 29 37 46 56 67 79 92 106 130 146 163 181 200 220 241 263 286 310 335 361 388 416 512 1024 // the mapping from quantization values to dpcm codes (array index) xdata short Q_VALUES[64] = {0, // 0 -quant31, // 1 -quant30, // 2 -quant29, // 3 -quant28, // 4 -quant27, // 5 -quant26, // 6 -quant25, // 7 -quant24, // 8 -quant23, // 9 -quant22, // 10 -quant21, // 11 -quant20, // 12 -quant19, // 13 -quant18, // 14 -quant17, // 15 -quant16, // 16 negative middle -quant15, // 17 -quant14, // 18 -quant13, // 19 -quant12, // 20 -quant11, // 21 -quant10, // 22 -quant9, // 23 -quant8, // 24 -quant7, // 25 -quant6, // 26 -quant5, // 27 -quant4, // 28 -quant3, // 29 -quant2, // 30 -quant1, // 31 0, // 32 Rev. 0.2 49 AN278 quant1, quant2, quant3, quant4, quant5, quant6, quant7, quant8, quant9, quant10, quant11, quant12, quant13, quant14, quant15, quant16, quant17, quant18, quant19, quant20, quant21, quant22, quant23, quant24, quant25, quant26, quant27, quant28, quant29, quant30, quant31}; // // // // // // // // // // // // // // // // // // // // // // // // // // // // // // // 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 positive middle 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 //----------------------------------------------------------------------------// Function PROTOTYPES //----------------------------------------------------------------------------unsigned char DPCM_Encode (short sample_diff); short DPCM_Decode (unsigned char dpcm_code); //----------------------------------------------------------------------------// DPCM_Encode //----------------------------------------------------------------------------// // Return Value : // 1) char dpcm_code - the 6-bit quantized DPCM code // range is positive range of 6-bit value: 0 to 63 // Parameters : // 1) short sample_diff - the difference between the predicted value and // the sample from the ADC // range is: -4096 to 4095 (difference of 12-bit values) // // Encode the sample using DPCM compression. // // The coding uses the following scheme (0 is unused) for an 8-bit sample: // // code: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 // q value: -64 -32 -16 -8 -4 -2 -1 0 1 2 4 8 16 32 64 // // The difference will be rounded down if positive and rounded up if // negative (i.e. 41 => 32, and -41 => -32). // // NOTE: the calling function must have the same register context, so it must // either have the keyword "using 2" or all "using 2" keywords need to be // removed // 50 Rev. 0.2 AN278 unsigned char DPCM_Encode (short sample_diff) using 2 { short sample_diff_us; unsigned char dpcm_code; // determine if the difference is positive or negative if (sample_diff < 0) { sample_diff_us = -sample_diff; // use the absolute value } else { sample_diff_us = sample_diff; } // narrow down which bits need to be set to use the proper quantization code // using a binary search algorithm (divide in halves) // the sign of the difference no longer matters // first tier if (sample_diff_us >= quant16) { // second tier if (sample_diff_us >= quant24) { // third tier if (sample_diff_us >= quant28) { // fourth tier if (sample_diff_us >= quant30) { // fifth tier if (sample_diff_us >= quant31) { dpcm_code = 63; } // fifth tier else { dpcm_code = 62; } } // fourth tier else { // fifth tier if (sample_diff_us >= quant29) { dpcm_code = 61; } // fifth tier else { dpcm_code = 60; } } } // third tier else { // fourth tier if (sample_diff_us >= quant26) { // fifth tier Rev. 0.2 51 AN278 if (sample_diff_us >= quant27) { dpcm_code = 59; } // fifth tier else { dpcm_code = 58; } } // fourth tier else { // fifth tier if (sample_diff_us >= quant25) { dpcm_code = 57; } // fifth tier else { dpcm_code = 56; } } } } // second tier else { // third tier if (sample_diff_us >= quant20) { // fourth tier if (sample_diff_us >= quant22) { // fifth tier if (sample_diff_us >= quant23) { dpcm_code = 55; } // fifth tier else { dpcm_code = 54; } } // fourth tier else { // fifth tier if (sample_diff_us >= quant21) { dpcm_code = 53; } // fifth tier else { dpcm_code = 52; } } } // third tier else { 52 Rev. 0.2 AN278 } } // fourth tier if (sample_diff_us >= quant18) { // fifth tier if (sample_diff_us >= quant19) { dpcm_code = 51; } // fifth tier else { dpcm_code = 50; } } // fourth tier else { // fifth tier if (sample_diff_us >= quant17) { dpcm_code = 49; } // fifth tier else { dpcm_code = 48; } } } // first tier else { // second tier if (sample_diff_us >= quant8) { // third tier if (sample_diff_us >= quant12) { // fourth tier if (sample_diff_us >= quant14) { // fifth tier if (sample_diff_us >= quant15) { dpcm_code = 47; } // fifth tier else { dpcm_code = 46; } } // fourth tier else { // fifth tier if (sample_diff_us >= quant13) { dpcm_code = 45; } // fifth tier else Rev. 0.2 53 AN278 { } dpcm_code = 44; } } // third tier else { // fourth tier if (sample_diff_us >= quant10) { // fifth tier if (sample_diff_us >= quant11) { dpcm_code = 43; } // fifth tier else { dpcm_code = 42; } } // fourth tier else { // fifth tier if (sample_diff_us >= quant9) { dpcm_code = 41; } // fifth tier else { dpcm_code = 40; } } } } // second tier else { // third tier if (sample_diff_us >= quant4) { // fourth tier if (sample_diff_us >= quant6) { // fifth tier if (sample_diff_us >= quant7) { dpcm_code = 39; } // fifth tier else { dpcm_code = 38; } } // fourth tier else { // fifth tier if (sample_diff_us >= quant5) { 54 Rev. 0.2 AN278 dpcm_code = 37; } // fifth tier else { dpcm_code = 36; } } // // // // if { } } } } } // third tier else { // fourth tier if (sample_diff_us >= quant2) { // fifth tier if (sample_diff_us >= quant3) { dpcm_code = 35; } // fifth tier else { dpcm_code = 34; } } // fourth tier else { // fifth tier if (sample_diff_us >= quant1) { dpcm_code = 33; } // fifth tier else { dpcm_code = 32; } } } convert the DPCM code to its 2's compliment if the original sample difference was negative For example, 41 (101001), which represents a difference of 60, 2's complimented becomes 23 (010111), which represents a difference of -60 (sample_diff < 0) dpcm_code = ~dpcm_code + 1; dpcm_code &= 0x3F; // use the 2's compliment of the dpcm // code // use only the 6 LSBs for the dpcm code return dpcm_code; //----------------------------------------------------------------------------// DPCM_Decode //----------------------------------------------------------------------------// // Return Value : // 1) short predicted_value - the signed and quantized difference between Rev. 0.2 55 AN278 // the predicted_value and the ADC sample, which is used // create the predicted_value for the next DPCM cycle // range is: -4096 to 4095 (difference of 12-bit values) // Parameters : // 1) char dpcm_code - the 6-bit code indicating the quantized difference // between the old_prediction and the current sample value // range is positive range of 6-bit value: 0 to 63 // // Decode the DPCM code to a signed difference between the current predicted // value and the next. // // NOTE: the calling function must have the same register context, so it must // either have the keyword "using 2" or all "using 2" keywords need to be // removed short DPCM_Decode (unsigned char dpcm_code) using 2 { return Q_VALUES[dpcm_code]; } //----------------------------------------------------------------------------// End Of File //----------------------------------------------------------------------------- 56 Rev. 0.2 AN278 LED Functions //----------------------------------------------------------------------------// F411_VR_LED.c //----------------------------------------------------------------------------// Copyright 2006 Silicon Laboratories, Inc. // http://www.silabs.com // // Program Description: // // This file contains the functions that use the PWM to brighten, dim, and // flutter the LEDs. // // These functions work by using a pointer (LED_DCH) to an "LED byte," which // is just a byte in memory associated with each LED. In F411_VR.c, // one of the Timer ISRs updates the PCA PWM registers with all the LED bytes // every interrupt period, so the LEDs that don't change are still updated, // but visually nothing changes. // // When the timer interrupts, PCA0CPH1 is reloaded with the current value // LED0_DC, which is changed by the functions (Dim, Brighten, and Flutter) // based on the desired LED behavior. By decrementing the time the LED is // on in steps (based on the ADJ variable), the LED appears to "dim" off, // and by incrementing the time the LED is on in steps, the LED appears to // "brighten" slowly. // // The LED_DCH pointer must be pointed to the correct LEDx_DC byte BEFORE // each of these functions is called. // // LED_DCH -> LED0_DC -> PCA0CPH1, where CEX1 (output from the PCA) is tied // to LED0 // // For example, the resulting dim LED waveform might look something like // this, since the LEDs are ON when CEX1 = 0: // // _ __ ___ // CEX1 _____________________________| |____________| |___________| | // // | 1st step | 2nd step | 3rd step | 4th step | // // (continued) // ____ _____ ______ _______ // CEX1 __________| |_________| |________| |_______| | // // | 5th step | 6th step | 7th step | 8th step | // // (continued) // ________ _________ __________ ___________ // CEX1 ______| |_____| |____| |___| | // // | 9th period | 10th period | 11th period | 12th period | // // (continued) // ____________ _____________________________________________ // CEX1 __| |_| // // | 13th period | 14th period | 15th period | 16th period | // // The LED has appeared to "dim" slowly off. // // // // NOTE: The calling function must have the same register context as the LED // functions, so it must either have the keyword "using 0" or all "using 0" Rev. 0.2 57 AN278 // // // // // // // // // // // // // // // // // // // // // // // // // // // // // // // // // // keywords for the LED functions need to be removed. How To Use: FID: Target: Tool chain: Project Name: See Readme.txt 41X000008 C8051F411 Keil C51 7.50 / Keil EVAL C51 Silicon Laboratories IDE version 2.6 F411_VR Release 1.4 -All changes by TP -06 Feb 2009 -No changes to this file Release 1.3 -All changes by TP -02 Feb 2006 -project version updated, no changes to this file Release 1.2 -All changes by TP -21 Nov 2005 -project version updated, no changes to this file Release 1.1 -All changes by TP -16 Aug 2004 -project version updated, no changes to this file Release 1.0 -Initial Revision (TP) -15 AUG 2004 //----------------------------------------------------------------------------// Includes //----------------------------------------------------------------------------#include <c8051f410.h> // SFR declarations //----------------------------------------------------------------------------// Global Variables //----------------------------------------------------------------------------unsigned char ADJ = 15; unsigned int LED_PWM = 65535; int LED_PWM_CHANGE = 0x0000; unsigned char *LED_DCH; unsigned char LED0_DC = 0x00; unsigned char LED1_DC = 0x00; // add another LEDx_DC variable here, if desired, and point to it with *LED_DCH // before calling the LED functions //----------------------------------------------------------------------------// Function PROTOTYPES //----------------------------------------------------------------------------void Dim_LED (void); void Brighten_LED (void); void Flutter_LED (void); //----------------------------------------------------------------------------// Dim_LED //----------------------------------------------------------------------------// 58 Rev. 0.2 AN278 // Return Value : None // Parameters : None // // Dim the LED using the PCA in 8-bit PWM mode. The Timer0 ISR in // F411_VR.c updates the value LED_DCH is pointing to. // // NOTE: This function requires that the LED_DCH pointer be "pointing" to the // appropriate LED byte, as explained above. // void Dim_LED (void) using 0 { // retrieve the previous value of the duty cycle unsigned char duty_cycle = *LED_DCH; ADJ = 0xF1; LED_PWM = 65535; LED_PWM_CHANGE = 0; TCON |= 0x10; // // // // // set the ADJ such that the LED will get dimmer reset the Timer 0 interval do not change the Timer 0 interval start Timer 0 // wait until the LED is fully off while (duty_cycle != 0x00) { duty_cycle = *LED_DCH; } TCON &= ~0x10; } // stop Timer 0 (no more updates to the // PCA duty cycle) //----------------------------------------------------------------------------// Brighten_LED //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // Brighten the LED using the PCA in 8-bit PWM mode. The Timer0 ISR in // F411_VR.c updates the value LED_DCH is pointing to. // // NOTE: This function requires that the LED_DCH pointer be "pointing" to the // appropriate LED byte, as explained above. // void Brighten_LED (void) using 0 { // retrieve the previous value of the duty cycle unsigned char duty_cycle = *LED_DCH; ADJ = 0x0F; LED_PWM = 65535; LED_PWM_CHANGE = 0; TCON |= 0x10; // // // // // set the ADJ such that the LED will brighten reset the Timer 0 interval do not change the Timer 0 interval start Timer 0 // wait until the LED is fully on while (duty_cycle != 0xFF) { duty_cycle = *LED_DCH; } TCON &= ~0x10; } // stop Timer 0 (no more updates to the // PCA duty cycle) Rev. 0.2 59 AN278 //----------------------------------------------------------------------------// Flutter_LED //----------------------------------------------------------------------------// // Return Value : None // Parameters : None // // Cause the LED to dim on and off. The Timer0 ISR in F411_VR.c updates // the value LED_DCH is pointing to. // // NOTE: This function requires that the LED_DCH pointer be "pointing" to the // appropriate LED byte, as explained above. // void Flutter_LED (void) using 0 { // retrieve the previous value of the duty cycle unsigned char duty_cycle = *LED_DCH; // check if the LED is currently on or off if (duty_cycle == 0xFF) { ADJ = 0xF1; } else { ADJ = 0x0F; } LED_PWM = 65535; // reset the Timer 0 interval LED_PWM_CHANGE = -200; // change the Timer 0 interval each // interrupt cycle so the LED has a // "fluttering" effect TCON |= 0x10; // start Timer 0 // Wait for a flutter cycle to finish while (LED_PWM > 17000) { } TCON &= ~0x10; } // stop Timer 0 (no more updates to the // PCA duty cycle) //----------------------------------------------------------------------------// End Of File //----------------------------------------------------------------------------- 60 Rev. 0.2 AN278 NOTES: Rev. 0.2 61 AN278 CONTACT INFORMATION Silicon Laboratories Inc. 400 West Cesar Chavez Austin, TX 78701 Tel: 1+(512) 416-8500 Fax: 1+(512) 416-9669 Toll Free: 1+(877) 444-3032 Please visit the Silicon Labs Technical Support web page: https://www.silabs.com/support/pages/contacttechnicalsupport.aspx and register to submit a technical support request. Patent Notice Silicon Labs invests in research and development to help our customers differentiate in the market with innovative low-power, small size, analogintensive mixed-signal solutions. Silicon Labs' extensive patent portfolio is a testament to our unique approach and world-class engineering team. The information in this document is believed to be accurate in all respects at the time of publication but is subject to change without notice. Silicon Laboratories assumes no responsibility for errors and omissions, and disclaims responsibility for any consequences resulting from the use of information included herein. Additionally, Silicon Laboratories assumes no responsibility for the functioning of undescribed features or parameters. Silicon Laboratories reserves the right to make changes without further notice. Silicon Laboratories makes no warranty, representation or guarantee regarding the suitability of its products for any particular purpose, nor does Silicon Laboratories assume any liability arising out of the application or use of any product or circuit, and specifically disclaims any and all liability, including without limitation consequential or incidental damages. Silicon Laboratories products are not designed, intended, or authorized for use in applications intended to support or sustain life, or for any other application in which the failure of the Silicon Laboratories product could create a situation where personal injury or death may occur. Should Buyer purchase or use Silicon Laboratories products for any such unintended or unauthorized application, Buyer shall indemnify and hold Silicon Laboratories harmless against all claims and damages. Silicon Laboratories and Silicon Labs are trademarks of Silicon Laboratories Inc. Other products or brandnames mentioned herein are trademarks or registered trademarks of their respective holders. 62 Rev. 0.2