DsPIC30F 5011 Development Board

From OpenCircuits
(Difference between revisions)
Jump to: navigation, search
(adding todo)
Line 2,100: Line 2,100:
*program the bootloader into flash under linux platform

Revision as of 20:06, 3 July 2008



Features of dsPIC30F5011

  • 2.5 to 5V
  • Up to 30MIPs
  • High current/sink source I/O pins: 25mA
  • DSP Instruction Set
  • Dual programming techniques: ICSP and RTSP
  • UART: up to 2 modules
  • I2C: up to 1Mbps
  • 10-bit A/D, 1.1 Msps
  • 12-bit A/D, 200 ksps
  • 44K flash (66Kb), 4Kb RAM, 1Kb EEPROM
  • No DAC
  • Pin-to-pin compatible with other dsPICs
Table 1.1 Comparison with Compatible dsPICs
dsPic Price
MIPs Flash
IC OC Motor
30F5011 5.91 30 66 4 1 52 16 8 8 0 5x16bit
0 2 2 1 2 1
30F6011A 7.73 30 132 6 2 52 16 8 8 0 5x16bit
0 2 2 1 2 0
30F6012A 7.85 30 144 8 4 52 16 8 8 0 5x16bit
0 2 2 1 2 1
33FJ128GP206 4.62 40 128 8 0 53 18 8 8 0 9x16bit
0 2 2 1 0 1
33FJ128GP306 4.81 40 128 16 0 53 18 8 8 0 9x16bit
0 2 2 2 0 1
33FJ128GP706 5.49 40 128 16 0 53 18 8 8 0 9x16bit
0 2 2 2 2 1
33FJ128MC506 4.97 40 128 8 0 53 16 8 8 8 9x16bit
1 2 2 2 1 0
33FJ128MC706 5.38 40 128 16 0 53 16 8 8 8 9x16bit
1 2 2 2 1 0
33FJ256GP506 6.11 40 256 16 0 53 18 8 8 0 9x16bit
0 2 2 2 1 1

Web Page



Code Examples

Programming Methods

  • There are 2 programming methods: In-Circuit Serial Programming (ICSP) and Run-Time Self-Programming (RTSP)
  • ICSP allows the devices to be programmed after being placed in a circuit board.
  • RTSP allows the devices to be programmed when an embedded program is already in operation.

ICSP: External Programmer (ICD2)

  • Two types of ICSP are available: ICSP and Enhanced ICSP. Both of them require setting MCLR# to VIHH (9V – 13.25V).
  • Standard ICSP
    • Use external programmer (e.g. MPLAB® ICD 2, MPLAB® PM3 or PRO MATE® II) only.
    • Required low-level programming to erase, program and verify the chip.
    • Slower, because codes are serially executed.
    • Program memory can be erased using Normal-Voltage (4.5 – 5.5V) or Low-Voltage (2.5V – 4.5V).
  • Enhanced ICSP
    • Use external programmer and Programming Executive (PE).
    • PE is stored in the on-chip memory.
    • PE allows faster programming.
    • PE can be downloaded to the chip by external programmer using the standard ICSP method.
    • PE contains a small command set to erase, program and verify the chip, avoiding the need of low-level programming.

Hardware Interface

Table 2.1 Pin Used by ICSP
Pin Label Function Pin Number
MCLR# Programming Enable 7
VDD Power Supply 10, 26, 38, 57
VSS Ground 9, 25, 41, 56
PGC Serial Clock 17
PGD Serial Data 18

Table 2.2 Available Programmers in the Market
Product Name Interface with PC Interface with Device Price (US) Postage (US) Total (US)
MPLAB® ICD 2 USB or RS232 6-PIN RJ-12 connector $159.99 - -
Full Speed USB Microchip ICD2
Debugger and Programmer
USB 6-PIN ICSP connector
6-PIN RJ-12 connector
$72.00 $12.00 $84.00
Mini Microchip Compatible ICD2
Debugger and Programmer
RS232 6-PIN ICSP connector
6-PIN RJ-12 connector
$45.00 $10.00 $55.00
ICDX30 RS232 6-pin RJ-11 $51.00 $47.46 $98.46
*Clone Microchip ICD2 (Now Using) USB 6-pin flat cables $30.00 $12.00 $42.00

Table 2.3 DIY ICD 2 Programmer Circuit
Source Schematic PIC16F877A Bootloader
Patrick Touzet Yes HEX
Nebadje Yes Zip

Software Interface

  • The program can be written and compiled in an Integrated Development Environment (IDE) using either Assembly or C. The complied codes are then loaded to the device through the external programmer.

Table 2.4 Summary of IDE
Product Name Features OS Price (US$)
MPLAB® IDE Assembler Only Windows Free
MPLAB® C30 Assembler and C-Compiler Windows $895.00 (Free student version1)
Piklab 0.12.0 Assembler and C-Compiler Linux Free2
  1. Full-featured for the first 60 days. After 60 days, some code optimization functions are disabled. The compiler will continue to function after 60 days, but code size may increase.
  2. The current version supports external programmer ICD 2, but not yet tested.

RTSP: COM Port (Bootloader)

  • RTSP works in normal voltage (MCLR# no need to raise to VIHH).
  • No literature has mentioned the incorporation of Programming Executive (PE). Presumably, since Enhanced ICSP needs to set MCLR# to VIHH, RTSP cannot use PE.
  • Refer to bootloader section.

IC Requirements

Table 3.1 IC Requirements
Part No. Description Min Temp Max Temp Min Volt Max Volt Typ Cur Max Cur
dsPIC30F5011-30I/PT uP -40oC 85oC 2.5V [1] 5.5V 145mA 217mA
MAX3232ESE RS232 driver -40oC 85oC 3.0V 5.5V 0.3mA 1.0mA
DS3695N RS485 driver -40oC 85oC 4.75V 5.25V 42mA 60mA
DAC6574IDGS 10-bit Quad-DAC I2C -40oC 105oC 2.7V 5.5V 0.6mA 0.9mA
74HC14D Quad-Schmitt Trigger -40oC 125oC 2.0V 6.0V 0.02mA
Overall -40oC 85oC 4.75V 5.25V <300mA [2]
dsPIC33FJ128GP306-I/PT uP -40oC 85oC 3.0V [1] 3.6V 74mA 250mA
ADM3485EARZ RS485 driver -40oC 85oC 3.0V 3.6V 1.1mA 2.2mA
24LC256-I/SN 256kBits I2C EEPROM -40oC 85oC 2.5V 5.5V 400uA 3mA
LM3940IMP-3.3 5V-3.3V Regulator -40oC 125oC 5.0V 7.5V 10mA 250mA
  1. Minimum voltage measured is 3.3V (with 2 LEDs blinking) running at 30MHz.
  2. Measured current at 5V is 180mA (with 2 LEDs blinking only)

Development Environment


PIC setup win.JPG

  • C-Compiler, Assembler and Linker are under GNU license.
    • MPLAB C30 C Compiler (*.c -> *.s)
    • MPLAB ASM30 Assembler (*.s -> *.o)
    • MPLAB LINK30 Linker (*.o -> *.bin)
  • PA optimizer, simulator, runtime libraries, header files, include files, and linker scripts are not covered by GNU. Reference is here.
  • Microchip has integrated ASM30, LINK30, assembly header files, linker scripts in MPLAB IDE, which is free for download.
  • MPLAB C30 costs US$895. A 60-day free student version is also available. After 60-days, the optimizer is automatically disabled, while other tools can still function properly. Refer to Table 2.4.

Table 4.1 C Libraries in MPLAB C30
Library Directory
(\\Microchip\MPLAB C30)
Major functions
DSP Library
(e.g. libdsp-coff.a)
Vector, Matrix, Filter, etc.
16-Bit Peripheral Libraries
(e.g. libp30F5011-coff.a)
ADC12, IOPort, UART, I2C, etc.
Standard C Libraries
(e.g. libc-coff.a, libm-coff.a, libpic-coff.a)
stdio.h, time.h, float.h, math.h,
MPLAB C30 Built-in Functions none _buildin_addab, _buildin_add, _buildinmpy, etc


PIC setup linux.JPG

  • C Compiler, Assembler and Linker are under GNU license.
    • The code can be downloaded from Microchip at here.
    • Current MPLAB ASM30 Assembler: v2.04
    • Current MPLAB C30 Compiler: v2.04
  • John Steele Scott has made templates that can be readily used by Debian-based systems.
  • For v1.32, the necessary conversion to *.deb has been done already at here.
    • Download pic30-1.32-debian.tar.bz2 for Template v1.32.
    • Download pic30-binutils_1.32-1_i386.deb for the assember.
    • Download pic30-gcc_1.32-1_i386.deb for the compiler.
  • For v2.00
  • For v3.01, convert the Toolchain following instructions at here
    • Pre-install these packages: dpkg-dev, debhelper, bison, flex, sysutils, gcc-3.3, fakeroot
      • cmd: sudo apt-get install dpkg-dev debhelper bison flex sysutils gcc-3.3 fakeroot
    • Download and unzip template: pic30-3.01.tar.bz2
    • Download assembler: mplabalc30v3_01_A.tar.gz. Save under /pic30-3.01/pic30-binutils-3.01/upstream/
    • Download c-compiler: mplabc30v3_01_A.tgz. Save under /pic30-3.01/pic30-gcc-3.01/upstream/
    • Install MPLAB_C30_v3_01-StudentEdition under Windows
    • Copy directories /include, /lib, /support, and /bin/c30_device.info to pic30-3.01/pic30-support-3.01/upstream/
    • Pack pic30-binutils into deb file
      • goto /pic30-3.01/pic30-binutils-3.01/
      • type cmd: dpkg-buildpackage -rfakeroot -b
    • Pack pic30-gcc-3.01 into deb file
      • goto /pic30-3.01/pic30-gcc-3.01/
      • type cmd: dpkg-buildpackage -rfakeroot -b
    • Pack pic30-gcc-3.01 into deb file
      • goto /pic30-3.01/pic30-support-3.01/
      • type cmd: dpkg-buildpackage -rfakeroot -b
    • install pic30-binutils_3.01-1_i386.deb
      • type cmd: sudo dpkg -i pic30-binutils_3.01-1_i386.deb
    • install pic30-gcc_3.01-1_i386.deb
      • type cmd: sudo dpkg -i pic30-gcc_3.01-1_i386.deb
    • install pic30-support_3.01-1_all.deb
      • type cmd: sudo dpkg -i pic30-support_3.01-1_all.deb
  • Important Note: Only the compiler is free. The header files and library are owned by Microchip.
    • Thomas Sailer suggested to download the Student version of C30 compiler and then build the libraries without source code. A package template for Fedora system is available here.
    • Instructions for filling the upstream direction is available here.
    • Alteratively, Stephan Walter has started a project to develop C Runtime Library for dsPIC.
      • Current libraries in version 0.1.1 include: assert.h, cdefs.h, ctype.h, errno.h, inttypes.h, stdint.h, stdio.h, stdlib.h, string.h
  • Burning Program Codes to Target Board
  1. Use 'dspicprg and dspicdmp' utilities developed by Homer Reid to burn hex code (*.hex) to devices. See Reference here. Through serial port only?
  2. Use Piklab IDE. Details on file format not known.
  3. Use MPLAB IDE to burn hex code (*.hex) to devices.

Code Optimization

  • Below is a comparsion between different optimization levels for the project including drivers for 2 projects.
Table 4.2 Comparison between differnt optimization levels
Optimization Description Project 1
Code Size
Project 1
Data Usage
Project 2
Code Size
Project 2
Data Usage
O0 No optimization
Fastest Compilation
6222 (9%) 178 (4%) 26,037 (38%) 710 (17%)
O1 Optimize
Tries to reduce code size and execution time.
4473 (6%) 178 (4%) 22,290 (32%) 710 (17%)
O2 Optimize even more
Performs nearly all supported optimizations
that do not involve a space-speed trade-off.
Increases both compilation time and the
performance of the generated code.
4422 (6%) 178 (4%) 21,993 (32%) 710 (17%)
O3 Optimize yet more.
O3 turns on all optimizations specified by O2
and also turns on the inline-functions option.
4485 (6%) 178 (4%) 22,176 (32%) 710 (17%)
Os Optimize for size.
Os enables all O2 optimizations that do not
typically increase code size. It also performs
further optimizations designed to reduce code
4356 (6%) 178 (4%) 21,885 (32%) 710 (17%)

Software Architecture

 Application   | Task 1 | Task 2 | Task 3 | Task 4 | Task 5 |
               |                 POSIX API                  |
    OS         |    Coroutine      |   FreeRTOS Scheduler   |
               |                   Drivers                  |
  Hardware     | UART | ADC | DAC | EEPROM |  PWM  | TIMERS | 
  • Currently, operating system is based on FreeRTOS incorporating coroutine developed by Simon Tatham
  • Software Drivers are to be developed to allow users at Application Level to use the hardware (e.g. ADC, DAC, UART, EEPROM etc) through the OS.
  • The interface between the drivers and the OS is based on POSIX standard (e.g. open(), write(), read(), ioctl(), usleep() etc).
  • The most up-to-date development can be found at repository freertos_posix

Programming Tips

Memory Map for 5011

Table 6.1 Memory Location
Type Start Address End Address Size
Flash 0x000000 0x00AFFF 44K[1]
+--Flash: Reset Vector 0x000000 0x000003 4
+--Flash: Interrupt Vector Table 0x000004 0x00007F 124
+--Flash: Alternate Vector Table 0x000084 0x0000FF 124
+--Flash: User Program 0x000100 0x00AFFF 43.7K
EEPROM 0x7FFC00 0x7FFFFF 1K[2]
Programming Executive 0x800000 0x8005BF 1472
Unit ID 0x8005C0 0x8005FF 64
Config Registers 0xF80000 0xF8000F 16
Device ID 0xFF0000 0xFF0003 4

[1] Each address is 16-bit wide. Every two addresses correspond to a 24-bit instruction. Each even address contains 2 valid bytes; each odd address contains 1 valid byte plus 1 phathom byte.
[2] Each address is 8-bit wide.

Data Location

Table 6.2 Data Location
Type Description Example
_XBSS(N) [1] RAM Data in X-memory, aligned at N, no initilization int _XBSS(32) xbuf[16];
_XDATA(N) [1] RAM Data in X-memory, aligned at N, with initilization int _XDATA(32) xbuf[] = {1, 2, 3, 4, 5};
_YBSS(N) [1] RAM Data in Y-memory, aligned at N, no initilization int _YBSS(32) ybuf[16];
_YDATA(N) [1] RAM Data in Y-memory, aligned at N, with initilization int _YDATA(32) ybuf[16] = {1, 2, 3, 4, 5};
__attribute__((space(const))) Flash ROM data, constant, accessed by normal C
statements, but 32K max.
int i __attribute__((space(const))) = 10;
__attribute__((space(prog))) Flash ROM data, read/write by program space visibility
window (psv)
int i __attribute__((space(prog)));
__attribute__((space(auto_psv))) Flash ROM data, read by normal C statements, write
by accessing psv
int i __attribute__((space(auto_psv)));
__attribute__((space(psv))) Flash ROM data, read/write by (psv) int i __attribute__((space(psv)));
_EEDATA(N) [1] ROM Data in EEPROM, aligned at N, read/write with psv int _EEDATA(2) table[]={0, 1, 2, 3, 5, 8};
_PERSISTENT RAM Data, data remain after reset int _PERSISTENT var1, var2;
_NEAR RAM Data at near section int _NEAR var1, var2;
_ISR Interrupt service rountine void _ISR _INT0Interrupt(void);
_ISRFAST Fast interrupt service rountine void _ISRFAST _T0Interrupt(void);
  1. N must be a power of two, with a minimum value of 2.

Configuration Bits

  • System clock source can be provided by:
  1. Primary oscillator (OSC1, OSC2)
  2. Secondary oscillator (SOSCO and SOSCI) with 32kHz crystal
  3. Internal Fast RC (FRC) oscillator at 7.37MHz (7372800Hz)
  4. Low-Power RC (LPRC) oscillator (Watchdog Timer) at 512 kHz.
  • These clock sources can be incorporated with interal Phase-locked-loop (PLL) x4, x8 or x16 to yield the osciallator frequrence FOSC
  • The system clock is divided by 4 to yield the internal instruction cycle clock, FCY=FOSC/4
  • FRC with PLLx16 is used to achieve FCY=29.49MHz (29491200Hz or 30MIPS)
 //The code (MACRO) below is to be placed at the top of program (before main)
    _FWDT(WDT_OFF);    //Turn off Watchdog Timer
    _FGS(CODE_PROT_OFF); //Disable Code Protection


  • Each timer is 16-bit (i.e. counting from 0 to 65535).
  • Timer 2 and 3 can be incorporated together to form a 32-bit timer.
  • Prescale is the ratio between timer counts and system clock counts. Prescales of 1:1, 1:8, 1:64 and 1:256 are available.
  • Timers may be used to implement free time clock or mesaure time.

Free Time Clock

  • Let required time for ticking be PERIOD.
  • Number of instruction cycles during PERIOD = PERIOD*FCY cycles
  • Using a prescale of 1:x, the timer period count register = # of cycles/x
  • e.g. PERIOD = 10ms; # of cycles = 10ms*30MHz = 300000 cylces; Using 1:64 Prescale, register setting = 300000/64 = 4688
  void time_init(void){
     TMR1 = 0;		// Clear register
     PR1 = 4688;	// Set period
     _T1IF = 0;		// Clear interrupt flag
     _T1IE = 1;		// Enable interrupts
     T1CONbits.TCS = 0;		// Use internal clock source
     T1CONbits.TCKPS = 2;	// Prescale Select 1:64
     T1CONbits.TON = 1;		// Start the timer 
  void _ISRFAST _T1Interrupt(void){
     _T1IF = 0;		// Clear interrupt flag
     //Place user code here

Time Measurement

  • To measure the time taken for action(), use the code below:
  unsigned int measure_time(void){	
     PR3 = 0xFFFF;		// Set counter to maximum
     _T3IF = 0;			// Clear interrupt flag
     _T3IE = 0;			// Disable interrupt
     T3CONbits.TON = 1;		// Start the timer, TMR3 count up
     TMR3 = 0;			//Clear TMR3 to start count up
     //Add code here to wait for something to happen
     T3CONbits.TON = 0;	//Stop the timer
     return (unsigned int) TMR3/FCY;      //TMR/FCY yields the actual time


  • Registers are involved in Interrupts includes:
  1. Interrupt Flag Status (IFS0-IFS2) registers
  2. Interrupt Enable Control (IEC0-IEC2) registers
  3. Interrupt Priority Control (IPC0-IPC10) registers
  4. Interrupt Priority Level (IPL) register
  5. Global Interrupt Control (INTCON1, INTCON2) registers
  6. Interrupt vector (INTTREG) register
  • User may assign priority level 0-7 to a specific interrupt using IPC. Setting priority to 0 disable a specific interrupt. Level 7 interrupt has the highest priority.
  • Current priority level is stored in bit<7:5> of Status Register (SR). Setting Interrupt Priority Level (IPL) to 7 disables all interrupts (except traps).
  • sti() and cli() can be defined to enable and disable global interrupts for time critical functions:
 #define IPL              ( 0x00e0 )
 #define cli()            SR |= IPL    //Set IPL to 7
 #define sti()            SR &= ~IPL   //Set IPL to 0
 char adc_ioctl(unsigned char request, unsigned char* argp){
   cli();		//Disable global interrupt
      adc_add_ch(argp[ch]);	//Add adc channels
   sti();		//Enable global interrupt
   return 0;
  • dsPic30F has an errate note on the Interrupt Controller. When Nested Interrupt is turned on (NSTDIS=0 by default), a high priority interrupt negating a low priority interrupt may result in an Address Error.
  • To work around the problem, it is suggested by Microchip to use the following MACRO to protect:
  1. the clearing of Interrput Flag
  2. the disabling of Interrupt Enable
  3. the lowering of Interrupt Priority
  4. the modification of IPL in Status Register to 1-6
  #define DISI_PROTECT(X)        { \
                                     __asm__ volatile ("DISI #0x1FFF");\
                                     X; \
                                     DISICNT = 0; \
  • For example,
  void _ISR _T1Interrupt( void )
      DISI_PROTECT(IFS0bits.T1IF = 0);
      //do something here...


  • 5011 provides two UART channels UxART, for x=1, 2.
  • UxMODE, UxSTA, UxBRG are registers used to set the mode, indicate the status, and set the baud rate respectively.
  • For UART communications compatiable with RS232 standard, an external driver (e.g. MAX3232ESE) is needed.
  • For UART communications compatiable with RS485 standard, an external driver (e.g. DS3695N) is needed.

Auto baud rate detection

  • The method is provided by ingenia bootloader.
  • The PC sends a ASCII character 'U' (0x55) to the target board.
  • On the first rising edge of the start bit, the target board starts the timer.
  • At the fifth rising edge, the timer is stopped, let the count number be t_count.
    • The measured period corresponds to 8 bits transmitted at a baud rate uxbrg.
    _   _   _   _   _   _
  _|S|_|1|_|1|_|1|_|1|_|S|_  (S = Start Bit)
   Measured Time
  • The relationship between uxbrg and TMR is
  Measured Time (in seconds) = t_count/Fcy
  uxbrg = 1/(Measured Time/8)
        = 8*Fcy/t_count
  • Since UxBRG is computed by:
  UxBRG = (Fcy/(16*Baudrate)) -1
        = (Fcy/(16*8*Fcy/t_count)) -1
        = t_count/128 -1
  • The following is the code for auto baud rate detection for U2ART:
  unsigned int uart2_autobaud(void){
     U2MODEbits.ABAUD = 1;		//Enable Autobaud detect from U2RX (from IC2 if 0)
     U2MODEbits.UARTEN = 1;		//U2ART enable
     //Timer 3 Config==========================================================
     PR3 = 0xFFFF;			// Set counter to maximum
     _T3IF = 0;			// Clear interrupt flag
     _T3IE = 0;			// Disable interrupt
     T3CONbits.TON = 1;		// Start the timer, TMR3 count up
     //Input Capture Config====================================================
     IC2CONbits.ICM = 3;		//Detect rising	
     _IC2IF = 0;			//Clear interrupt flag
     _IC2IE = 0; 			//Disable interrupt
     //Start Auto baud detection===============================================
     unsigned int i=0;
     cli();				//Disable Global Interrupt
     while(!_IC2IF);			//1st rising edge detected
     TMR3 = 0;			//Clear TMR3 to start count up
     _IC2IF = 0;			//Clear interrupt flag
     while(!_IC2IF);			//2nd rising edge detected
     _IC2IF = 0;			//Clear interrupt flag
     while(!_IC2IF);			//3rd rising edge detected
     _IC2IF = 0;			//Clear interrupt flag
     while(!_IC2IF);			//4th rising edge detected
     _IC2IF = 0;			//Clear interrupt flag
     while(!_IC2IF);			//5th rising edge detected
     _IC2IF = 0;			//Clear interrupt flag
     T3CONbits.TON = 0;		//Stop the timer
     sti(); 			//Enable Global Interrupt
     //Compute value for BRG register==========================================
     unsigned int time;
     time = ((TMR3+0x40)>>7)-1;	//+0x40 for rounding
     return time;
  • For 30MIP, tested speeds of transmission include 9600bps, 19200bps, 28800bps, 38400bps and 57600bps.


  • The following structures and variables are used as circular buffers for transmit and receive.
  struct UART_Rx{
      unsigned char   wr;
      unsigned char   rd;
  struct UART_Tx{
      unsigned char   wr;                             
      unsigned char   rd;
      unsigned char   tx_complete_flag;
  struct UART_Rx uart_rx;
  struct UART_Tx uart_tx;
  unsigned char uart_rx_buf[MAX_UART_RX_BUF];
  unsigned char uart_tx_buf[MAX_UART_TX_BUF];
  char uart_open()
     uart_rx.wr = 0;
     uart_rx.rd = 0;
     uart_tx.wr = 0;
     uart_tx.rd = 0;
     uart_tx.tx_complete_flag = 1;
     return 0;
  void uart2_init(void){
     unsigned int u2brg = 97; 
     u2brg = uart2_autobaud();
     U2BRG  = u2brg;					
     // Disable U2ART
     U2MODEbits.UARTEN = 0;			//Disable U2ART module
     // Configure Interrupt Priority
     _U2RXIF = 0;	//Clear Rx interrupt flags
     _U2TXIF = 0;	//Clear Tx interrupt flags
     _U2RXIE = 1;	//Receive interrupt: 0 disable, 1 enable 
     _U2TXIE = 1;	//Transmit interrupt: 0 disable, 1 enable
     // Configure Mode
     //  +--Default: 8N1, no loopback, no wake in sleep mode, continue in idle mode
     //  +--Diable autobaud detect
     //  +--Enable U2ART module
     U2MODEbits.ABAUD = 0;	//Disable Autobaud detect from U2RX 
     U2MODEbits.UARTEN = 1;	//U2ART enable
     // Configure Status
     //  +--Default: TxInt when a char is transmitted, no break char
     //  +--Default: RxInt when a char is received, no address detect, clear overflow
     //  +--Enable Transmit
     U2STAbits.UTXEN = 1;	//Tx enable


  • This function writes a series of bytes to the circular buffer and start transmission.
  int uart_write(unsigned char *buf, int count)
     //If transimt has not completed, return busy
     if(uart_tx.tx_complete_flag == 0){
        return -1;            
        uart_tx.tx_complete_flag = 0;
     int next_data_pos;
     int byte = 0;
     for (; byte<count; byte++) {
        next_data_pos = pre_wr_cir254buf(   (unsigned char)uart_tx.wr, 
                                            (unsigned char)uart_tx.rd, 
        if (next_data_pos!=255) {
            //Valid data is available
            uart_tx_buf[uart_tx.wr] = (unsigned char) buf[byte];    //copy the char to tx_buf
            uart_tx.wr = next_data_pos;                                     //increment the ptr
        } else break;
     //Raise Interrupt flag to initiate transmission
     _U2TXIF = 1;    //Start interrupt
     return byte;        
  • The interrupt routine reads from the circular buffer and send the data. The uart is opened such that the module will generate an TX Interrupt when it a byte is sent.
  void _ISR _U2TXInterrupt(void){
     DISI_PROTECT(_U2TXIF = 0);		//Clear Interrupt Flag
     unsigned char next_data_pos;
     next_data_pos = pre_rd_cir254buf( (unsigned char)uart_tx.wr,
                                       (unsigned char)uart_tx.rd, 
     if (next_data_pos!= 255) {
        //Valid Data is available to transmit
        U2TXREG = (uart_tx_buf[(unsigned char)uart_tx.rd] & 0xFF);  //send next byte...
        uart_tx.rd = (unsigned char) next_data_pos;    //update rd pointer
     } else {
        //Transimission has completed
       uart_tx.tx_complete_flag = 1;    // change to empty of tx


  • The interrupt routine writes to the circular buffer when space is available.
  void _ISR _U2RXInterrupt(void){
     unsigned char next_data_pos;
     if ( U2STAbits.URXDA ){
        next_data_pos = pre_wr_cir254buf( uart_rx.wr, uart_rx.rd, MAX_UART_RX_BUF);
        if (next_data_pos!=255) {
           //If buffer is not full
           uart_rx_buf[uart_rx.wr] = (unsigned char) U2RXREG; //Read the data from buffer
           uart_rx.wr = next_data_pos;
           //When buffer is full, still remove data from register, butthe incoming data is lost
           next_data_pos = (unsigned char) U2RXREG; 			//Read the data from buffer
     DISI_PROTECT(_U2RXIF = 0);        //Clear the flag
  • This function reads one byte from the circular buffer.
  int uart_read(unsigned char *buf)
     int next_data_pos;
     next_data_pos = pre_rd_cir254buf( uart_rx.wr, uart_rx.rd, MAX_UART_RX_BUF);
     //Copy 1 byte when data is available
     if (next_data_pos!=255) 
        *buf = uart_rx_buf[uart_rx.rd];     //copy the stored data to buf
        uart_rx.rd = next_data_pos;                 //update the ptr
        return 1;
     //No data can be copied
         return 0;


  • Two lines are devoted for the serial communication. SCL for clock, SDA for data.
  • Standard communication speed includes
  1. Standard speed mode: 100kHz
  2. Fast speed mode: 400kHz
  3. High speed mode: 3.4MHz
  • dsPIC30f5011 supports standard and fast speed modes. The maximum speed attainable is 1MHz.
  • Pull-up resistors are required for both SCL and SDA. Minimum pull-up resistance is given by:
    Pull-up resistor (min) = (Vdd-0.4)/0.003  ......  [See section 21.8 in Family reference manual]
  • 2.2Kohm is typical for standard speed mode.
  • After initiating a start/stop/restart bit, add a small delay (e.g. no operation) before polling the corresponding control bit (hardware controlled).
  • After sending a byte and receiving an acknowledgement from the slave device, ensure to change to idle state.


  • The following structure is used to record whether special bits are needed to be sent.
  typedef union{
      unsigned char val;
          unsigned START:1;       //start
          unsigned RESTART:1;     //restart
          unsigned STOP:1;        //stop
          unsigned NACK:1;        //not acknowledgment
          unsigned :1;
          unsigned :1;
          unsigned :1;
          unsigned :1;    
  static I2C_STATUS i2c_status;   
  • Initializing I2C with default speed I2C_BRG without interrupts.
  void i2c_open(void)
      //Open i2c if not already opened
      if(I2CCONbits.I2CEN == 0)
          _SI2CIF = 0;        //Clear Slave interrupt
          _MI2CIF = 0;        //Clear Master interrupt
          _SI2CIE = 0;        //Disable Slave interrupt
          _MI2CIE = 0;        //Disable Master interrupt
          I2CBRG = I2C_BRG;
          I2CCONbits.I2CEN = 1;   //Enable I2C module 
          i2cIdle();              //I2C bus at idle state, awaiting transimission
          i2c_status.val = 0;     //clear status flags


  • Use this function before read/write to append special bits before or after the data byte.
  char i2c_ioctl(unsigned char request, unsigned char* argp)
          case I2C_SET_STATUS:
              i2c_status.val = *argp;
              return -1;      //request code not recognised   
      return 0;


  • This function sends an 8-bit data using the I2C protocol.
  Mst/Slv    _______ M ____M___ S M ________ 
  SDA (Data)        |S|  data  |A|S|
                    |T|        |C|T|
  • Use ioctl() to select whether a start/restart/stop bit is required.
  • If slave does not respond after ACK_TIMEOUT, the transmission is considered unsucessful.
  int i2c_write(unsigned char *buf)
      unsigned int count = 0;
          I2CCONbits.SEN = 1;                 
          Nop();                          //A small delay for hardware to respond
          while(I2CCONbits.SEN);          //Wait till Start sequence is completed
      else if(i2c_status.bits.RESTART)
          I2CCONbits.RSEN = 1;                
          Nop();                          //A small delay for hardware to respond
          while(I2CCONbits.RSEN);         //Wait till Start sequence is completed
      I2CTRN = *buf;                  //Transmit register
      while(I2CSTATbits.TBF);         //Wait for transmit buffer to empty
          if(++count > ACK_TIMEOUT){
              //Slave did not acknowledge, byte did not transmit sucessfully, 
              //send stop bit to reset i2c
              I2CCONbits.PEN = 1;
              Nop();                          //A small delay for hardware to respond
              while(I2CCONbits.PEN);          //Wait till stop sequence is completed
              return 0;
          I2CCONbits.PEN = 1;
          Nop();                          //A small delay for hardware to respond
          while(I2CCONbits.PEN);          //Wait till stop sequence is completed
      i2c_status.val = 0;             //Clear status
      return 1;


  • This function reads 1 byte from slave using the I2C protocol.
  Mst/Slv     ____ ___S____ M M _____    
  SDA (Data)      |  data  |A|S|
                  |        |C|T|
  • Use ioctl() to select whether an ACK/NACK and/or STOP bit is needed to be sent.
  int i2c_read(unsigned char *buf)
      I2CCONbits.RCEN = 1;                    //Enable Receive
      I2CSTATbits.I2COV = 0;                  //Clear receive overflow
      *buf = (unsigned char) I2CRCV;          //Access the receive buffer
      I2CCONbits.ACKDT = (i2c_status.bits.NACK)? 1 : 0;
      I2CCONbits.ACKEN = 1;       //Send Acknowledgement/Not Acknowledgement
      i2cIdle();                  //I2C bus at idle state, awaiting transimission
          I2CCONbits.PEN = 1;
          Nop();                          //A small delay for hardware to respond
          while(I2CCONbits.PEN);          //Wait till stop sequence is completed
      i2c_status.val = 0;             //Clear status
      return 1;


  Mst/Slv    _______ M ___M___ M S ____M___ S M ___M___ M S ___S____ M ___S____ M M _____ 
  SDA (Data)        |S|       | |A|        |A|R|       | |A|        |A|        |N|S|
                    |T|address|W|C|channelA|C|E|address|R|C| Data H |C| Data L |A|T|
   * Send start bit, slave address (Write Mode)
  status = I2C_START;
  i2c_ioctl(I2C_SET_STATUS, &status);
  data = (unsigned char) I2C_SLAVE_ADDR;
   * Send control byte: Channel select
  data = (unsigned char) ctrl_byte;
   * Send restart bit, slave address (Read Mode)
  status = I2C_RESTART;
  i2c_ioctl(I2C_SET_STATUS, &status);
  data = (unsigned char) (I2C_SLAVE_ADDR|0x01);
   * Receive High Byte with Acknowledgment
  usr_data.high = (unsigned char) data;
   * Receive Low Byte with Not Acknowledgment and stop bit
  status = I2C_NACK | I2C_STOP;
  i2c_ioctl(I2C_SET_STATUS, &status);
  usr_data.low = (unsigned char) data;


  • 12-bit ADC: (Max 16 Channels)
  • Allow a maximum of 2 sets of analog input multiplexer configurations, MUX A and MUX B (Normally use one only).
  • A maximum of 200kps of sampling rate when using auto sampling mode.


  • The following variables are required.
  unsigned int adc_buf[ADC_MAX_CH];   //Store most updated data
  volatile unsigned int* ADC16Ptr = &ADCBUF0; //Pointer to ADC register buffer, 
  unsigned char adc_ch_select = 0;   //Pointer to channel to be read from
  unsigned char adc_data_ready = 0;   //Indicate if RAM data is ready for output
  • Configuration is highlighted below.
    • Interrupt: The ADC module will be set to interrupt when the specified channels are updated.
    • I/O: Set the corresponding TRISBX bits (digit i/o config) to input (i.e. = 1), and set corresponding bits in ADPCFG (analog config) to zero.
    • Scanning Mode: Scan mode is used. In this mode, the Sample and Hold (S/H) is switched between the channels specified by ADCSSL (Scan select register).
    • Reference Voltage for S/H: Only MUX A is used. By default, the negative reference voltage of the S/H is connected to VREF-.
    • Settings for ADC Operation: For 200kbps operation, the voltage references for the ADC voltage are connected to VREF+ and VREF-. Scan input is enabled, and the module will generate an interrupt when all selected channels have been scanned.
    • Sampling Rate: TAD refers to the time unit for the ADC clock. To configure the ADC module at 200kbps, the minimum sampling time of 1TAD = 334ns is required. ADCS<5:0> in ADCON3 register is used to set the time, which is given by:
     ADCS<5:0> = 2(TAD/TCY)-1 
               = 2(334e-9/33.34e-9)-1 
               = 19
  char adc_open(int flags)
     // Configure interrupt
     _ADIF = 0;                          //clear ADC interrupt flag
     _ADIE = 1;                          //enable adc interrupt
    // Configure analog i/o  
    _TRISB0 = 1;
    _TRISB1 = 1;    
    ADPCFG = 0xFFFC;                    //Enable AN0 (Vref+) and AN1 (Vref-)
    // Configure scan input channels    
    ADCSSL = 0x0003;    //0 => Skip, 1 => Scan
    // Configure CH0 Sample and Hold for 200kbps
    //  +-- Use MUX A only
    //  +-- Set CH0 S/H -ve to VRef-
    ADCHSbits.CH0NA = 0;
    // ADCCON3:
    //  +--Auto Sample Time = 1TAD
    //  +--A/D Conversion Clock Source = system clock
    //  +--A/D Conversion Clock Select ADCS<5:0>= 2(TAD/TCY)-1
    //      200kbps(Sampling frequency)
    ADCON3bits.SAMC = ADC_ACQ_TIME;     //1TAD for sampling time
    ADCON3bits.ADRC = 0;                //Use system clock
    ADCON3bits.ADCS = ADC_ADCS;         //each conversion requires 14TAD
    // ADCCON2:
    //  +--Default: Use MUX A, No splitting of Buffer
    //  +--Voltage Reference Configuration Vref+ and Vref-
    //  +--Scan Input Selections
    //  +--5 samples between interrupt
    ADCON2bits.VCFG = 3;                //External Vref+, Vref-
    ADCON2bits.CSCNA = 1;               //Scan input
    ADCON2bits.SMPI = 1;                //take 2 samples (one sample per channel) per interrupt
    // ADCCON1:
    //  +--Default: continue in idle mode, integer format
    //  +--Enable ADC, Conversion Trigger Source Auto, Auto sampling on
    ADCON1bits.FORM = 0;                //[0:integer]; [2 fractional]; [3 siged fractional]
    ADCON1bits.SSRC = 7;                //auto covert, using internal clock source
    ADCON1bits.ASAM = 1;                //auto setting of SAMP bit
    ADCON1bits.ADON = 1;                //Turn on module
    return 0;


  • 16 registers (ADCBUF0 -ADCBUF15) are dedicated to store the ADC data between interrupts. However, the data in ADCBUFx does not necessarily correspond to the data taken for channel x. Since the lowest register will always be filled first, when some of the channels are not scanned (i.e. skipped), care must be taken. The following code checks the ADCSSL register for the current scanning channels and moves the data to the corresponding position in *adc_buf.
  void _ISR _ADCInterrupt(void){
     unsigned int channel = 0;
     unsigned int buffer = 0;
     for (; channel<ADC_MAX_CH; channel++)
        if(select(channel))       //Check if channel has been selected
           adc_buf[channel] = ADC16Ptr[buffer];     //Copy data to adc_buf
     adc_data_ready = 1;
     DISI_PROTECT(_ADIF = 0);  //Clear adc interrupt
  static unsigned char select(unsigned char ch)
      unsigned int mask;
      mask = 0x0001 << ch;
      if(ADCSSL & mask)
          return 1;
      return 0;
  • User can read from the buffer at anytime to get the most updated analog values.
  int adc_read(unsigned int* buf, int count)
     if(adc_data_ready == 1)
        int num_channel = count/2;                  //number of channels to read
        unsigned char channel = adc_ch_select;     //index for adc_buf
        int i = 0;                                  //index for buf
        while(i<num_channel && channel<ADC_MAX_CH)
           //Loop only for specified number of channel or all channels 
           buf[i++] = adc_buf[channel++];      //use data in local buffer
           {  //increment to next valid channel
              if(channel >= ADC_MAX_CH) break;
       return 2*i;
     return -1;


  • This function is used to add or remove channels from the ADC scanning process.
  char adc_ioctl(unsigned char request, unsigned char* argp)
          case ADC_ADD_CH:
              //ADD channels to current set==========================
              cli();                      //Disable global interrupt
              if(select(argp[0]) == 0){   //If channel not in scan list
                  adcAdd(argp[0]);            //Add individual channel to scan list
                  adc_data_ready = 0;         //First data not ready yet, until interrupt occurs
              adc_ch_select = argp[0];    //Select current channel for reading
              sti();                      //Enable global interrupt
          case ADC_RM_CH:
              //REMOVE channels from current set==========================
              cli();                  //Disable global interrupt
              if(select(argp[0])){    //If channel in scan list       
                  adcRm(argp[0]);             //Remove individual channel
                  adc_ch_select = 0;          //Reset to AN0
              sti();                  //Enable global interrupt
              return -1;      //request code not recognised   
      return 0;
  • Channels may be added or removed by changing _TRISBX, ADPCFG, ADCSSL and ADCON2bits.SMPI.
  void adc_add_ch(unsigned char ch){
     unsigned int mask;
     mask = 0x0001 << ch;
     TRISB = TRISB | mask;
     ADCSSL = ADCSSL | mask;		 
     ADCON2bits.SMPI++;	//take one more sample per interrupt
  void adc_rm_ch(unsigned char ch){
     unsigned int mask;
     mask = 0x0001 << ch;
     ADPCFG = ADPCFG | mask;
     ADCON2bits.SMPI--;	//take one less sample per interrupt


  • 5011 has 1024 bytes of EEPROM, readable and writable under normal voltage (5V).
  • To use, declare:
 unsigned char _EEDATA(2) eeData[1024]={ 0x00, 0x00, 0x00, 0x00, .... }
 unsigned int byte_pointer = 0;


  • This function moves the pointer to the desired position before a reading/writing operation is performed.
  int eeprom_lseek(int offset, unsigned char whence){
     byte_pointer = offset;
     return byte_pointer;


  • This function read count bytes from the eeprom.
  int eeprom_read(unsigned char* buf, int count){
     int i=0;
     for(; i<count && byte_pointer < 1024; i++){
        readEEByte( __builtin_tblpage(eeData), 
                    __builtin_tbloffset(eeData) + byte_pointer, 
        byte_pointer++;		//Update global pointer
     return i;	//read i bytes successful	
  • readEEByte() is implemented in assembly code as follows:
  .global _readEEByte
     push      TBLPAG		;w0 = base of eeData
     mov       w0, TBLPAG	;w1 = offset for eeData in byte
     tblrdl.b  [w1], [w2]	;w2 = pointer to user buffer
     pop     	TBLPAG


  • This function write count bytes to eeprom.
  int eeprom_write(unsigned char* buf, int count){
     char isOddAddr = byte_pointer%2;	//current address is odd
     char isOddByte = count%2;		//number of bytes to write is odd
     unsigned int word_offset = byte_pointer>>1; //div by 2 and round down
     int max_write;
     max_write = (isOddAddr == 0 && isOddByte == 0) ? (count>>1) : (count>>1)+1;
     unsigned int word_data;	//Store word to be written
     int byte_wr = 0;		//number of bytes written, i.e buffer pointer
     int i = 0;
     for(; i<max_write && word_offset<512; i++, word_offset++){
        if(i==0 && isOddAddr){
           //First byte not used
           //============================================save first byte
           readEEByte( __builtin_tblpage(eeData), 
                       __builtin_tbloffset(eeData) + byte_pointer - 1,
           word_data = ((unsigned int)buf[0] << 8) + (0xFF & word_data);
           byte_wr++;				//Update buffer pointer
           byte_pointer++;			//Update global pointer
         } else if(i==max_write-1 && ((isOddAddr && sOddByte==0)||(isOddAddr==0 && isOddByte))){
           //Last byte not used
           //=============================================save last byte
           readEEByte(	__builtin_tblpage(eeData), 
                       __builtin_tbloffset(eeData) + byte_pointer + 1,
           word_data = (word_data << 8) + buf[byte_wr];
           byte_wr++;				//Update buffer pointer
           byte_pointer++;			//Update global pointer
         } else{
           //Both bytes valid
           word_data = ((unsigned int)buf[byte_wr+1] << 8) + buf[byte_wr];
           byte_wr+=2;				//Update buffer pointer
           byte_pointer+=2;		//Update global pointer
      eraseEEWord( __builtin_tblpage(eeData), 
                   __builtin_tbloffset(eeData) + 2*word_offset);
      writeEEWord( __builtin_tblpage(eeData), 
                   __builtin_tbloffset(eeData) + 2*word_offset,
      return byte_wr;		//No. of byte written
  • eraseEEWord and writeEEWord are implemented in assembly.
  .global _eraseEEWord
     push   TBLPAG				
     mov    w0, NVMADRU	;w0 = base of eeData
     mov    w1, NVMADR		;w1 = offset for eeData in word
     mov    #0x4044, w0			
     mov    w0, NVMCON		;Set to erase operation
     push   SR			;Disable global interrupts
     mov    #0x00E0, w0
     ior    SR
     mov    #0x55, w0 		;Write the KEY sequence
     mov    w0, NVMKEY
     mov    #0xAA, w0			
     mov    w0, NVMKEY
     bset   NVMCON, #15	;Start the erase cycle, bit 15 = WR
 L1: btsc   NVMCON, #15	;while(NVMCONbits.WR)
     bra    L1
     clr    w0
     pop    SR			;Enable global interrupts
     pop    TBLPAG
 .global _writeEEWord
     push   TBLPAG		;w0 = base of eeData
     mov    w0, TBLPAG		;w1 = offset for eeData in byte
     tblwtl [w2], [w1]		;w2 = pointer to user buffer
     mov    #0x4004, w0        ;Set to write operation
     MOV    w0, NVMCON
     push   SR			;Disable global interrupts
     mov    #0x00E0, w0
     ior    SR
     mov    #0x55, w0 		;Write the KEY sequence
     mov    w0, NVMKEY
     mov    #0xAA, w0			
     mov    w0, NVMKEY
     bset   NVMCON, #15	;Start the erase cycle, bit 15 = WR
 L2: btsc   NVMCON, #15	;while(NVMCONbits.WR)
     bra    L2
     clr    w0
     pop    SR			;Enable global interrupts
     pop    TBLPAG

Simple PWM (Output Compare Module)

  • The PWM module consists of 8 channels using the output compare module of dsPic.
  • These channels are locate at pin 46 (OC1), 49 (OC2), 50 (OC3), 51 (OC4), 52 (OC5), 53 (OC6), 54 (OC7), 55 (OC8). These pins are shared with port D.
  • The range of PWM freqeuencies obtainable is 2Hz to 15MHz (See Figure 6.3). Suggested range of operation is 2Hz to 120kHz. The relationship between resolution r and PWM frequency fPWM is given by:
        fPWM = fCY/(Prescale*10rlog(2))
Table 6.3 Relationship of Resolution and PWM Frequency
Resolution (bit) Prescale=1 Prescale=8 Prescale=64 Prescale=256
1 15,000,000 1,875,000 234,375 58,594
2 7,500,000 937,500 117,188 29,297
3 3,750,000 468,750 58,594 14,648
4 1,875,000 234,375 29,297 7,324
5 937,500 117,188 14,648 3,662
6 468,750 58,594 7,324 1,831
7 234,375 29,297 3,662 916
8 117,188 14,648 1,831 458
9 58,594 7,324 916 229
10 29,297 3,662 458 114
11 14,648 1,831 229 57
12 7,324 916 114 29
13 3,662 458 57 14
14 1,831 229 29 7
15 916 114 14 4
16 458 57 7 2


  • A timer (either Timer 2 or 3) is needed to determine the pwm period. The following code uses timer 2 for all 8 channels.
 void pwm_open(void){
   OC1CON = 0;	OC2CON = 0; //Disable all output compare modules
   OC3CON = 0; OC4CON = 0;
   OC5CON = 0; OC6CON = 0;
   OC7CON = 0; OC8CON = 0;
   TMR2 = 0;		// Clear register
   PR2 = 0xFFFF;	// Set to Maximum
   _T2IF = 0;		// Clear interrupt flag
   _T2IE = 0		// Enable interrupts
   T2CONbits.TCS = 0;		// Use internal clock source
   T2CONbits.TCKPS = 0;	// Prescale Select 1:1
   T2CONbits.TON = 1;		// Start the timer 	


  • User should select the channel and set the pwm period using the functions below before issuing the duty cycle:
 char pwm_ioctl(unsigned char request, unsigned long* argp){
   unsigned int value;
   unsigned char mask;
     case PWM_SET_PERIOD:
       return setPeriodNPrescale(argp[0]);
     case PWM_SELECT_CH:
       pwm_channel = argp[0];
       mask = 0x01 << pwm_channel;
       pwm_status = pwm_status | mask;
           return 0;
           return -1;
 char setPeriodNPrescale(unsigned long value_ns){
   unsigned long ans;
   unsigned long long numerator = (unsigned long long)value_ns*SYSTEM_FREQ_MHZ;
   int index= -1;
   unsigned long denominator;
       denominator = (unsigned long)1000*pwm_prescale[++index];
       ans = (unsigned long)(((long double)numerator/denominator) + 0.5) - 1; //round to nearest int
   } while(ans > 0x0000FFFF && index < 3);
   if(ans > 0x0000FFFF)
       return -1;
   T2CONbits.TON = 0;		// Turn off the timer
   T2CONbits.TCKPS = index;	// Change prescale factor
   PR2 = (unsigned int) ans;	// Set to Maximum
   T2CONbits.TON = 1;		// Turn on the timer 	
   return 0;


  • User can change the duty cycle using the following functions
 int pwm_write(unsigned long* buf){
   if((pwm_status & (0x01 << pwm_channel)) == 0){
       return -1;		//Channel has not been enabled
       case 0:
           OC1RS = calcDCycle(buf[0]); OC1R = OC1RS; 
           OC1CONbits.OCM = 6; //Simple PWM, Fault pin disabled
       case 1:
           OC2RS = calcDCycle(buf[0]);	OC2R = OC2RS; 
           OC2CONbits.OCM = 6; //Simple PWM, Fault pin disabled
       case 7:
           OC8RS = calcDCycle(buf[0]); OC8R = OC8RS; 
           OC8CONbits.OCM = 6; //Simple PWM, Fault pin disabled
           return -1;
   return 4;
 unsigned int calcDCycle(unsigned long value_ns){
   unsigned long long numerator = (unsigned long long)value_ns*SYSTEM_FREQ_MHZ;
   unsigned int index = T2CONbits.TCKPS;
   unsigned long denominator = (unsigned long)1000*pwm_prescale[index];
   return (unsigned int)(((long double)numerator/denominator) + 0.5) - 1; //round to nearest int

Propagration Delay

  • PWM channels sharing the same timer will have their PWM signals synchronised (i.e. the HIGH state of the duty cycle are all triggered together).
  • To introduced delay to the PWM signals, the signal from selected channels may be made to pass through a series of inverters (e.g. 74HC14D). This adds propagation delay to the signal.
  • However, as propagration delay of logic gates depends on applied voltage, temperature and load capacitance, the accuracy is low and performance is poor. For accurate delay, delay lines may be used, but they are expensive.
Table 6.4 Propagation Delay of Philips 74HC14D[1], [2]
3.3V 5.0V
Number of Gates A B C A B C
2 21ns (10.5) 23ns (11.5) 22ns (11.0) 15ns (7.5) 14ns (7.0) 14ns (7.0)
4 45ns (11.3) 46ns (11.5) 46ns (11.5) 30ns (7.5) 30ns (7.5) 30ns (7.5)
6 69ns (11.5) 70ns (11.7) 72ns (12.0) 45ns (7.5) 46ns (7.7) 47ns (7.8)

[1] Data in specification for 4.5V: Typical 15ns, Maximum 25ns
[2] Data in specification for 6.0V: Typical 12ns, Maximum 21ns

DSP Library

  • Library functions in <dsp.h> include the following categories:
  1. Vector
  2. Window
  3. Matrix
  4. Filtering
  5. Transform
  6. Control

Data Types

  • Signed Fractional Value (1.15 data format)
    • Inputs and outputs of the dsp functions adopt 1.15 data format, which consumes 16 bits to represent values between -1 to 1-2-15 inclusive.
    • Bit<15> is a signed bit, positive = 0, negative = 1.
    • Bit<14:0> are the exponent bits e.
    • Positive value = 1 - 2-15*(32768 - e)
    • Negative value = 0 - 2-15*(32768 - e)
  • 40-bit Accumulator operations (9.31 data format)
    • The dsp functions use the 40 bits accumalators during arithmatic calculations.
    • Bit<39:31> are signed bits, positive = 0x000, negative = 0x1FF.
    • Bit<30:0> are exponent bits.
  • IEEE Floating Point Values
    • Fractional values can be converted to Floating point values using: fo = Fract2Float(fr); for fr = [-1, 1-2-15]
    • Floating point values can be converted to Fractional values using: fr = Float2Fract(fo); or fr = Q15(fo); for fo = [-1, 1-2-15]
    • Float2Fract() is same as Q15(), except having saturation control. When +ve >= 1, answer = 215-1 = 32767 (0x7FFF). When -ve < -1, answer = -215 = -32767 (0x8000)

Build-in Library

  • Some assembler operators can only be accessed by inline assembly code, for example,
  1. Manuipulation of accumulators A and B (add, sub, mul, divide, shift, clear, square)
  2. Bit toggling
  3. Access to psv (program space visiblity) page and offset
  4. Access to table instruction page and offset
  • Built-in functions are written as C-like function calls to utilize these assembler operators.

Bootloader Development


  • Programming with ICSP is useful when the target board is produced in batch. The producer can download a program even when the chip is on the target board.
  • However, ICSP requires an external programmer.
  • To allow the user to change the program after production but without the need of an external programmer, bootloader becomes useful.
  • Bootloader is a small program installed via ICSP. Everytime the device is reset, the bootloader is run first. The bootloader first detects the default serial channel whether the user wishes to download a new program to the device. If so, the bootloader will pause there, and wait for the user to download the hex file from the PC. The hex file is written to the device via RTSP instructions in the bootloader. If a new download is not necessary, the bootloader redirects to the previously installed user's program.
  • The disadvantage of bootloaders is that they consume some of the memory of the device.

Table 7.1 Free bootloaders for dsPIC
Developer Source Platform User Guide Remarks
ingenia Assembly Windows pdf
  • Works for all dsPIC supporting RTSP
  • Auto baudrate detection
  • Use about 1.15% of the flash memory space (0xAFFF-0xAE00)/(0xAFFF-0x0100)
  • Development of Linux platform is underway
  • Modification of code for dsPIC30F5011 is successful
Tiny Assembly Windows Web
  • By default, only supports 601X, 601X, 401X, 2010
  • Smaller code size than ingenia, but not as easy to modify
Elektronika Hex Windows txt
  • Only works for dsPIC30F6014 serial port UART2 at baudrate 57600


  • The bootloader developed by ingenia is open source and it has been modified (see below) to suit our development using dsPic30f5011.
  • The bootloader (hereafter called dsPicBootloader) employs the following settings:
  1. Use U2ART channel
  2. Use FRC, PLL16
  3. For 5011, the bootloader is located between 0x00AE00 to 0x00AFFE (512bytes). Refer to C:\Program Files\Ingenia\ingeniadsPICbootloader\ibl_dspiclist.xml after installing the GUI interface.

1. including p30f5011.gld and p30f5011.inc

       .include "p30f5011.inc"

2. changing the config code of UART #0x8420 -> #0x8020

       ; Uart init
       mov #0x8020, W0           ; W0 = 0x8020 -> 1000 0000 0010 0000b
       mov W0, U2MODE            ; Enable UART, AutoBaud and 8N1
       clr U2STA

3. changing the start address 0xAE00 - 0x0100 = 0AD00

         .equ CRC, W4
         .equ ACK, 0x55
         .equ NACK, 0xFF
         .equ USER_ADDRESS, 0x0100
         .equ START_ADDRESS, 0xAD00                ; Relative to 0x0100

4. using Internal FRC and PLL16

       config __FOSC, CSW_FSCM_OFF & FRC_PLL16 ;Turn off clock switching and
                                          	;fail-safe clock monitoring and
                                           	;use the Internal Clock as the
                                          	;system clock

5. disabling MCLR (optional)

       config __FBORPOR, PBOR_ON & BORV_27 & PWRT_16 & MCLR_DIS
                                           ;Set Brown-out Reset voltage and
                                           ;and set Power-up Timer to 16msecs

6. changing all the related registers of U1ART to U2ART, all U1XXX => U2XXX


7. changing all the related registers of IC1 to IC2, all IC1XXX => IC2XXX

       IC2CON, #IC2IF, #IC2IE

dsPicProgrammer (Java-based Multi-Platformed)

  • Ingenia developed a programmer (PC-side) that works only in Windows environment. The project for Linux environment is currently suspended.
  • A simple programmer (hereafter called dsPicProgrammer) written in Java based on the library developed by RXTX has been developed here. The programmer supports both Linux and Windows environments, and may be used as a substitution for the official programmer developed by ingenia.
  • The programmer has the following specification and limitations:
  1. Can be used on both Linux and Windows platforms.
  2. Adjustable baudrate (9600bps to 57600bps[1]).
  3. Support programming of dsPIC30F5011 and dsPIC33FJ128GP306[2] devices (Developers may add your devices).
  4. Protection against overwriting bootloader codes on devices.
  5. Detection if application program does not have its reset() at user's code start address.
  6. Reprogramming can be done without powering down the target board, provided the user's program is compliant to that stated below.
  7. Target board will run the user's program after programming is done.
  8. Can be used with USB-Serial Cables. Below is a list of tested cable:
Prolific PL-2303 USB to Serial Bridge Controller: Driver download

[1] from version 1.5.2 onwards, the baudrate is increased to 115200bps.
[2] only dsPIC30 devices are compatible with the ingenia's programmer. For dsPIC33 devices, the protocol has been modified due to the increased flash size, and the dsPIC33 bootloader can only work with dsPicProgrammer.

Special Consideration

  • The bootloader assumes that the user program starts at address 0x100. This is usually the case, but there are always exceptions.
  • To ensure that the user program always starts at address 0x100, you can create a customized linker script and customized reset() function as follows:
  • Copy and modify the file named "crt0.s" from the directory "C:\Program Files\Microchip\MPLAB C30\src\pic30" to the project directory and include it.
   .section .reset, code      //previously .section .libc, code 
  • Copy and modify the linkerscript for the device (e.g. p30f5011.gld) to the project directory and include it.
  .text __CODE_BASE :
     *(.reset);              //<-insert this line here
     *(.libc) *(.libm) *(.libdsp);  /* keep together in this order */
  } >program

Communication Protocol

          +-------------------+             +-------------------+----------------+
          |  dsPicProgrammer  |             |  dsPicBootloader  | User's Program |
          +-------------------+             +-------------------+----------------+
          |       PC          |             |             Target dsPic           |
          +-------------------+             +------------------------------------+
          |     COM PORT      |=============|                 UART               |
          +-------------------+             +------------------------------------+
  • Stage 1: User's Configuation
    • Select a COM port channel
    • Select a baudrate
    • Select the user hex file
         java -jar dsPicProgrammer.jar COM1 19200 foo.hex
  • Stage 2: Resetting Target Device
    • dsPicProgrammer sends a Break character (pull UART-TX to low logic, which is normally high).
    • User's program on dsPic detects the break character and reset the chip
NOTE: The user's program is expected to have the following code in order to enable this function. Otherwise, the target board must be restarted manually.
  void _ISR _U2RXInterrupt(void)
      //No Framming error
      if( U2STAbits.FERR == 0)
          //Normal procedure
      //Framming error
          if ( U2STAbits.URXDA ){
              unsigned char data;
              data = (unsigned char) U2RXREG;
              if(data == 0x00){
                  // A break char has been received: 
                  //  U2RX has been pulled to zero for more than 13 bits
                  //  This is used to reboot the pic
                  mdelay(800);    //wait for break character to clear
                  asm("reset");   //software reset
      _U2RXIF = 0;        //Clear the flag
  • Stage 3: Entering Ingenia's Protocol
    • Transmission is conducted in 8N1, i.e. 8-bit, no parity, 1 stop-bit
    • Communication Protocol is reviewed in ingenia bootloader user's guide section 2.1.3. The following summarises the key steps on the PC side (Refer also to section 2.2.2).
  1. Autobaud rate detection: dsPicProgrammer continuously sends a character "U" [0x55] via COM port and waits for an acknowledgment character "U", [ACK] = [0x55]
  2. Version Control: dsPicProgrammer sends the command character [0x03]. On success, dsPicProgrammer receives 3 characters i) Major Version ii) Minor Version iii) Acknowledgment [0x55]
  3. Device ID Monitoring: dsPicProgrammer sends the read command character [0x01] + 24-bit address [High][Medium][Low] (0xFF0000). Then, it receives 4-byte data [High][Medium][Low][ACK]
  4. Load the user hex file and check integrity
  5. Start Programming: dsPicProgrammer issues the write command character [0x02] + 24-bit address [High][Medium][Low]+ Number of bytes [N] + [data 0] + [data 1] + ... + [data N-1] + [CRC]=(INTEL HEX8 Checksum - Sum modulo 256) and receives [ACK] or [NACK] = [0xFF]
NOTE: Writing is in row mode access (i.e. erase and write a whole row, each row has 32 instructions, or 96 bytes because each instruction has 24 bits)
  • Stage 4: Goto User's Program
    • dsPicProgrammer sends the goto user code command [0x0F]

USB-RS232 Bridge

  • As USB ports are becoming more and more common, COM ports and Parallel ports may be redundant in the next few years. This section explore the possibilities of programming the target board through a USB port.
  • There are two options:
  1. Use an external USB/RS232 adaptor, the driver will emulate a virtual COM port, such as Prolific and FDTI. Ingenia has tested its bootloader with some USB-232 manufacturers (silabs, FTDI, etc..). However, the programming failed with our Prolific adapter. Application program may use JavaComm API (javax.comm) and/or RXTX to drive the COM port.
  2. Modified the bootloader program on PC to support USB communication. e.g. using jUSB and JSR-80 (javax.usb). External circuits such as PIC18F4550 and MAX232 are required.
   |--User's App.--|-------Device Manager------|-------USB-RS232 Interface------|---dsPIC---|
  Option 1:
    +-------------+  +----------+  +----------+  +---+  +------------+  +-----+  +--------+
    | Application |--| JavaComm |--| Virtual  |==|USB|--|    FDTI    |--|RS232|==| Target |
    |   Program   |  |  RXTX    |  | COM Port |  +---+  | Circuitary |  +-----+  | Board  |
    +-------------+  +----------+  +----------+         +------------+           +--------+
  Option 2:
    +-------------+          +--------+          +---+  +------------+  +-----+  +--------+
    | Application |----------| JSR-80 |==========|USB|--| PIC18F4550 |--|RS232|==| Target |
    |   Program   |          |  jUSB  |          +---+  |   MAX232   |  +-----+  | Board  |
    +-------------+          +--------+                 +------------+           +--------+
  • Currently, when RXTX is incorporated with JavaComm API, operating systems supported include Linux, Windows, Mac OS, Solaris and other operating systems. On the other hand, jUSB and JSR-80 only works for linux.

FDTI Chipset

  • FT232RL communicates with PC via USB to provide 1 UART channel.
  • Datasheet can be downloaded here.
    • Refer to Fig. 11 (Page 19) for Bus Powered Configuration.
    • Refer to Fig. 16 (Page 24) for for UART TTL-level Receive [RXD -> 1], Transmit [TXD -> 4], Transmit Enable [CBUS2/TXDEN -> 3]. Omit Receive Enable [CBUS3/PWREN#] and use [CBUS2/TXDEN -> 2]
    • Refer to Fig. 15 (Page 23) for LED Configuration: [CBUS0/TXLED#] and [CBUS1/RXLED#]
  • Virtual COM Port Drivers can be downloaded here.

Programming the Device


  • Hardware
  1. PC with COM port (Windows XP Installed for MPLAB)
  2. ICD2 Programmer
  3. Target Board
  4. 5V Power Supply
  • Software
  1. MPLAB IDE v7.50: v7.60 is not compatible with our current ICD2 Programmers. There are consistent "Devices cannot be founded" warnings.
  2. dsPicProgrammer (dsPicProgrammer.jar)
  3. RXTX driver: download and upzip rxtx-2.1-7-bins-r2.zip (Final)
  • Files
  1. dsPicBootloader (bl_5011.hex). Original assembly code by ingenia can be downloaded from here.
  2. Application hex file (e.g. app.hex)

Loading Bootloader (Once only)

Table 9.1 Loading Bootloader
Step Remarks
  • Do NOT connect ICD 2 (via USB) to PC
  • Execute MPLAB vX.XX Install.exe
Install USB Driver
  • Follow the instruction in (C:\Program Files\Microchip\MPLAB IDE\ICD2\Drivers\Ddicd2.htm)
Select Target Chip
  • Run MPLAB IDE on PC
  • Select: Configure>Select Devices...
  • Choose dsPIC30F5011
Target <-> ICD 2
  • Use six pin cable. Beware of the pin assignments. Only pin 1 - 5 should be used.
  • Place Jumper on target board (if any). The Jumper connects target Vcc to ICD 2.
  • Do NOT power-up the target.
ICD 2 <-> PC
  • Plug-in ICD 2 to PC via USB cable
  • Power-up the target.
  • Select: Programmer>Select Programmer>MPLAB ICD 2
  • If this is the first time the ICD 2 is connected to PC, MPLAB IDE will automatically download the required OS to ICD 2, wait until it has finished
  • If you have not connected and powered up the target, you might see Warnings on invalid device IDs, and/or running self tests.
  • See results of self test if necessary: Programmer>Settings, Status Tab. Refer to ICD2 User's Guide Chapter 7.
Load Bootloader
  • Select: File>Import...
  • Select bl_5011.hex
Start Programming
  • Select: Programmer>Program
  • Power-down the Taget
  • Select: Programmer>Select Programmer>None
  • Unplug USB cable

Loading Firmware

Java Environment Setup

  • Download and install the latest JDK or JRE
  • Download and Extract RXTX Driver
    • Available from RXTX
    • File: rxtx-2.1-7-bins-r2.zip (Final)
    • Extract the files using software such as WinRAR
      1 2 rxtx.JPG
  • Copy RXTXcomm.jar
    • To C:\Program Files\Java\X\lib\ext (under the latest jre, e.g. X = jre1.6.0_03)
      1 3 rxtxcomm.JPG
  • Copy rxtxSerial.dll
    • To C:\Program Files\Java\X\bin (under the lastest jre, e.g. X = jre1.6.0_03)
    • For Linux users, copy librxtxSerial.so to /jre/lib/[machine type] (i386 for instance)
      1 4 rxtxserial.JPG
  • Download dsPicProgrammer
    • Available from here
    • Expand the latest tag under dsPicProgrammer
    • File: dsPicProgrammer.jar
      1 5 dspicprogrammer.JPG
    • Save the file (dsPicProgrammer.jar) and your hex file (foo.hex) to your local directory (e.g. C:\dsPicProgrammer\)
      1 6 dspicprogrammer2.JPG
directly download this package

Download Firmware

  • Start a Command Prompt
  • Run dsPicProgrammer
    • Make sure your device is connected through a "Direct RS232 Cable" to PC Serial Port.
    • Change to the directory where dsPicProgrammer.jar is located and run the command:
      • [path/to/java/]java -Djava.library.path=[path/to/rxtxlib] -jar dsPicProgrammer.jar COM1 57600 foo.hex
        • for example, jre\bin\java -Djava.library.path=.\rxtx\bin -jar dspicProgrammer.jar COM1 57600 foo.hex
      • Or directly using dsPicProgrammer.bat COM1 57600 foo.hex(Windows)
      • java -jar dsPicProgrammer.jar /dev/ttyS0 57600 foo.hex (Linux)
where COM1 is your COM Port ID
57600 is communication speed [in bps]
foo.hex is your firmware new file
2 3 command.JPG
    • When the program prompt you for an input, press y:
      File:2 4 prompt.JPG
    • If communication can be established, you should see something like this:
      2 5 progress.JPG
    • Wait until programming is completed.
      2 6 complete.JPG
    • after all completed, power OFF your device, then power ON again, enjoy your new updated.


  • Invalid COM Port
    • In case of selecting an invalid COM port, you should see the error message:
      gnu.io.NoSuchPortException: Choosing COM Port Error
      3 1 com err.JPG
  • Missing firmware file
    • In case of selecting a firmware file that does not exist, you should see the error message:
      java.io.FileNotFoundException: foo1.hex <The system cannot find the file specified>
      3 2 file err.JPG
  • Missing RXTX driver
    • In case of missing the RXTX driver, you should see the error message:
      Exception in thread “main” java.lang.NoClassDefFoundError: gnu/io/UnsupportedCommOperationException
      3 3 rxtx err.JPG

Remote Access

  • At the moment, local devices (e.g. EEPROM, ADC, DAC, etc.) can only be accessed locally through POSIX functions such as open(), read(), write(), ioctl().
  • However, a client may need to access these devices on a remote server. This section reviews the background and gives some ideas on its possible implementation.


  • A remote file access protocol, to transfer "files" (i.e. device's data) such as:
  1. File Transfer Protocol (FTP): Required files are copied from sever to client for manipulation
  2. Remote Shell (RSH): Required files are copied from sever to client for manipulation
  3. Network File System (NFS): Required files are manipulated on sever
  • An API to access files using a selected protocol, such as:
  1. lam_rfposix: A POSIX-like remote file service for Local Area Multicomputer
  2. API employed by VxWorks: VxWorks is a Unix-like real-time operating system, commonly used for embedded systems.

API Reference for VxWorks

Conversion to dsPIC33F Devices

  • This section discusses the conversion required from dsPIC30F5011 to dsPIC33FJ128GP306.
  • Refer to official document dsPIC30F to dsPIC33F Conversion Guidelines (DS70172A).
  • Note that this section does not mainly intend to introduce the new functionalities of dsPIC33F devices. It only serves the purpose to summarise the major (if not minimum) changes required to port the setup of dsPIC30 to dsPIC33 devices.


  • dsPIC33 operates at voltage of 3.3V. A voltage regulator, such as LM3940 can be used to convert 5V supply to 3.3V.
  • A 1uF capacitor has to be placed at pin 56 (previously VSS, now VDDCORE).


Configuration Bits

  • dsPIC33 can operate at 40MIPs at maximum. To configure the device using internal FRC, replace the configuration bits setting as follows:
  _FOSCSEL(FNOSC_FRCPLL);  // FRC Oscillator with PLL
                           // Clock Switching and Fail Safe Clock Monitor is disabled
                           // OSC2 Pin Function: OSC2(RC15) as Digital IO
                           // Primary Oscillator Mode: Disabled
  _FWDT(FWDTEN_OFF);       // Watchdog Timer Enabled/disabled by user software		
  • Configure on-chip PLL at runtime as follows (at start of main function):
  _PLLDIV = 38;            // M=40: PLL Feedback Divisor bits
  CLKDIV = 0;              // N1=2: PLL VCO Output Divider Select bits
                           // N2=2: PLL Phase Detector Input Divider bits
  OSCTUN = 22;              // Tune FRC oscillator, if FRC is used; 
                            // 0: Center frequency (7.37 MHz nominal)
                            // 22: +8.25% (7.98 MHz)
  RCONbits.SWDTEN = 0;      // Disable Watch Dog Timer
  while(OSCCONbits.LOCK != 1);	// Wait for PLL to lock


  • No change is required.


  • dsPIC33 supports upto 2 I2C devices. As a result, replace all I2C related registers with xxI2Cyy to xxI2C1yy. For examples:
  _SI2C1IF = 0; //Clear Slave interrupt
  _MI2C1IF = 0; //Clear Master interrupt
  _SI2C1IE = 0; //Disable Slave interrupt
  _MI2C1IE = 0; //Disable Master interrupt
  I2C1BRG = I2C_BRG; // Configure Baud rate
  I2C1CONbits.I2CEN = 1;


  • The ADC in dsPic33 is significantly different from that in dsPic30. Specifically, ADC in dsPic33 uses DMA to buffer the adc data. Replace the open, interrupt routine, add and remove codes as follows:
  unsigned int adc_bufA[ADC_MAX_CH] __attribute__((space(dma),aligned(256)));
  unsigned int adc_bufB[ADC_MAX_CH] __attribute__((space(dma),aligned(256)));
  unsigned int* ADC16Ptr;             //Pointer to ADC register buffer, 
  unsigned char adc_ch_select = 0;    //Pointer to channel to be read from
  unsigned char adc_data_ready = 0;   //Indicate if RAM data is ready for output
  unsigned int which_dma = 0;         //indicate which adc_buf to be used
  void adc_open(void)
      // Configure interrupt
      _AD1IF = 0;              //clear ADC interrupt flag
      _AD1IE = 0;              //disable adc interrupt
      AD1CHSbits.CH0NA = 0;
      // Configure analog i/o  
      _TRISB0 = 1;
      _TRISB1 = 1;    
      AD1PCFG = 0xFFFC;        //Enable AN0 (Vref+) and AN1 (Vref-)
      AD1PCFGH = 0xFFFF;       //AN16-AN31: Disabled
      // Configure scan input channels    
      AD1CSSL = 0x0003;    //0 => Skip, 1 => Scan
      AD1CSSH = 0x0000;    //Skipping AN16-AN31
      // ADCCON4:
      AD1CON4bits.DMABL = 0;    // Each buffer contains 1 word
      // ADCCON3:
      AD1CON3bits.SAMC = 1; //1TAD for sampling time
      AD1CON3bits.ADRC = 0;            //Use system clock
      AD1CON3bits.ADCS = ADC_ADCS;     //each conversion requires 14TAD
      // ADCCON2:
      AD1CON2bits.VCFG = 3;    //External Vref+, Vref-
      AD1CON2bits.CSCNA = 1;   //Scan input
      AD1CON2bits.SMPI = 1;    //2 channels are scanned
      // ADCCON1:
      AD1CON1bits.FORM = 0;        //[0:integer]; [2 fractional]; [3 siged fractional]
      AD1CON1bits.SSRC = 7;        //auto covert, using internal clock source
      AD1CON1bits.ASAM = 1;        //auto setting of SAMP bit
      AD1CON1bits.AD12B = 1;       //12-bit, 1-channel ADC operation
      AD1CON1bits.ADDMABM = 0;     // DMA buffers are built in scatter/gather mode
      AD1CON1bits.ADON = 1;        // Turn on the A/D converter
      // DMA0 Configuration:
      DMA0CONbits.AMODE = 2;      // Configure DMA for Peripheral indirect mode
      DMA0CONbits.MODE  = 2;      // Configure DMA for Continuous Ping-Pong mode
      DMA0CNT = 1;                // generate dma interrupt every 2 samples 
                                  // same as SMPI because only 1 dma buffer per channel         
      DMA0REQ = 13;               // Select ADC1 as DMA Request source
      DMA0STA = __builtin_dmaoffset(adc_bufA);     
      DMA0STB = __builtin_dmaoffset(adc_bufB);
      _DMA0IF = 0;                // Clear the DMA interrupt flag bit
      _DMA0IE = 1;                // Set the DMA interrupt enable bit
      DMA0CONbits.CHEN=1;         // Enable DMA
  void _ISR _DMA0Interrupt(void)
      ADC16Ptr = (which_dma == 0)? adc_bufA : adc_bufB;   //Update pointer
      adc_data_ready = 1;
      which_dma ^= 1;                                     //Next buffer to be used
      _DMA0IF = 0;                                        //Clear the DMA0 Interrupt Flag
  static void adcAdd(unsigned char ch){
      unsigned int mask;
      mask = 0x0001 << ch;
      TRISB = TRISB | mask;
      AD1CSSL = AD1CSSL | mask;      
      AD1PCFG = ~AD1CSSL;
      AD1CON2bits.SMPI++;  //take one more sample per interrupt
  static void adcRm(unsigned char ch){
      unsigned int mask;
      mask = 0x0001 << ch;
      AD1PCFG = AD1PCFG | mask;
      AD1CSSL = ~AD1PCFG;
      AD1CON2bits.SMPI--;  //take one less sample per interrupt


  • There is no EEPROM in dsPIC33 devices. Please consider to use an external EEPROM using I2C communication.

Simple PWM

  • No change is required.

Memory Map for dsPIC33FJ128GP306

Table 11.1 Memory Location
Type Start Address End Address Size
Flash 0x000000 0x0157FF 86K[1]
+--Flash: Reset Vector 0x000000 0x000003 4
+--Flash: Interrupt Vector Table 0x000004 0x0000FF 252
+--Flash: Alternate Vector Table 0x000104 0x0001FF 252
+--Flash: User Program 0x000200 0x0157FF 85.5K
Programming Executive 0x800000 0x800FFF 4K[1]
Config Registers 0xF80000 0xF80017 24
Device ID (0xE5) 0xFF0000 0xFF0003 4

[1] Each address is 16-bit wide. Every two addresses correspond to a 24-bit instruction. Each even address contains 2 valid bytes; each odd address contains 1 valid byte plus 1 phathom byte.

Custom Linker Script to Maximize Space for Constant Data

  • Constant data declared using keyword const will be stored in the .const section in the flash memory.
  • Normally, during compilation, the linker will assign these data after the program code (.text section).
  • Since .const is accessed by auto-psv function, to maximize the space for constant data (32kb), the .const section needs to be aligned at 0x80000 boundary.
  • This requires the following change in linker script:
 __CONST_BASE = 0x8000;
 .text __CODE_BASE :
       *(.libc) *(.libm) *(.libdsp);  /* keep together in this order */
       /* *(.text);		deleted to maximize space for const data */
 } >program
 .const __CONST_BASE :
 } >program
  • If your program is large, after this change in linker script, function calls may involve large jump in the memory map (>32kB). As a result, you may need to enable the large code and large memory model during compilation. In such case, use the following options in your build path:
   -mlarge-code -mlarge-data
  • Meanwhile, functions that are defined in the standard C libraries, but are replaced with your own implementations (e.g. I/O stubs: open(), read(), write(), lseek(), ioctl() etc.) may have the following linker error:
   /usr/pic30-elf/lib//libc-elf.a(fflush.eo)(.libc+0x3c): In function '.LM11':
   : Link Error: relocation truncated to fit: PC RELATIVE BRANCH _write
   /usr/pic30-elf/lib//libc-elf.a(fclose.eo)(.libc+0x42): In function '.LM18':
   : Link Error: relocation truncated to fit: PC RELATIVE BRANCH _close 
  • To resolve the problem, you need to place the functions in the .libc section rather than in the .text section, like this:
   #define LIBC_CODE_LOC __attribute__ ( (section(".libc")))
   int LIBC_CODE_LOC open(const char *pathname, int flags){ ... }
   int LIBC_CODE_LOC close(int fd){ ... }
   int LIBC_CODE_LOC write(int fd, void* buf, int count) { ... }
   int LIBC_CODE_LOC read(int fd, void* buf, int count) { ... }
   int LIBC_CODE_LOC ioctl(int fd, int request, void* argp) { ... }
   int LIBC_CODE_LOC lseek(int fd, int offset, int whence) { ... }

dsPicBootloader and dsPicProgrammer

  • RTSP for dsPIC33F is different from dsPIC30F.
    • Row size changes from 32 instructions (96bytes) to 64 instructions (192 bytes)
    • Erase operation changes from 1 row to 8 rows
    • No EEPROM
  • With regards to the above changes, dsPicBootloader and dsPicProgrammer has been modified. In particular, dsPicProgrammer can be used to program both dsPic30F and dsPic33F devices (Yet, dsPic33F is no longer compatible with Ingenia's programmer). You can easily add your devices to the source code.


Table 12.1 Related software download links for dsPicBootloader and dsPicProgrammer
Program Site 1 Site 2 Remarks
JDK website Download latest JDK
RXTX website Download rxtx-2.1-7-bins-r2.zip or later
dsPicBootloader click click (v1.3) Under "dsPicBootloader/", download bl_5011.s or bl_j128gp306.s
dsPicProgrammer click click (v1.3.5) Under "dsPicProgrammer/", dowload dsPicProgrammer.jar

Alternatively, if you want to compile yourself or modify the source code, download
all source files under "dsPicProgrammer/" plus RdFileIntelHex.java under
You should also install RXTX on your local machine as recommended in the readme file.
Ingenia's bootloader website Download original ingenia's bootloader


  • program the bootloader into flash under linux platform
Personal tools