Require specific Guide/Manual for Propeller 2 GPIO coding in c/c++

evanh

@chintan_joshi said:
oh Yes , thank you. with this update i am getting 25 c temperature. But after removing Temp_pin also i am getting same temperature so how this is possible?

Not sure exactly what removing means but an unconnected pin ADC will be at VIO/2. 25 C beta reference is also at VIO/2 (where Rt == R0).

That's why I used the 150 kR internal bias resistor in my testing. I didn't have a real thermistor so imagined one with the bias. EDIT: I probably should have called it a simulated thermistor rather than just a bland bias.

chintan_joshi

@evanh I have tried two different thermistors with P2. One is cartridge internal thermistor and another one is used for room temperature. Both thermistor showing temperature as 25 c. So looks like issue with our Analog read using P2.

Because one thermistor used for room temperature showing proper room temperature with Arduino interfacing.

Are we missing any setup for analogread using P2? Let me create new thread for Analog read using P2. So it will be proper reference for other developers.

evanh

More likely you're using the wrong pin. The fact that applying the bias resistor changes the reading also tells you the ADC is working.

chintan_joshi

@evanh i am using pin 14 as Temp_pin. Yes but with original Thermistor i am not getting proper values.

evanh

Hmm, hang-on, there is something weird going on when I change things. It might be a C optimiser issue ....

EDIT: Okay, no, that was me.

Separately, and more relevant, I've found that the quantity of decimations for settling weren't even close to enough. Dunno how it was working actually.

I've added a settling parameter to pinwait() and made it a single loop. My testing requires 10 or 11 decimations until settled - which is a lot! Must be something to do with the high input impedance. You may find a different value is better with a real thermistor. The test code tries settling values of 3 to 19.

PS: Settling of 11 means a longer 25 microseconds per sample.

chintan_joshi

@evanh said:
Hmm, hang-on, there is something weird going on when I change things. It might be a C optimiser issue ....

EDIT: Okay, no, that was me.

Separately, and more relevant, I've found that the quantity of decimations for settling weren't even close to enough. Dunno how it was working actually.

I've added a settling parameter to pinwait() and made it a single loop. My testing requires 10 or 11 decimations until settled - which is a lot! Must be something to do with the high input impedance. You may find a different value is better with a real thermistor. The test code tries settling values of 3 to 19.

PS: Settling of 11 means a longer 25 microseconds per sample.

Still its giving same output with thermistor.

( Entering terminal mode.  Press Ctrl-] or Ctrl-Z to exit. )                    

   clkfreq = 100000000   clkmode = 0x100090b                                    
      pin P14                                                                   
duration = 1433  gio = 11807  vio = 55176  value = 32625  Temperature = 22.91177
3  Ratio = 1.083245  rln = 0.079961                                             
duration = 1546  gio = 11818  vio = 55200  value = 33543  Temperature = 25.08248
8  Ratio = 0.996870  rln = -0.003135                                            
duration = 1618  gio = 11819  vio = 55212  value = 33553  Temperature = 25.09094
2  Ratio = 0.996549  rln = -0.003457                                            
duration = 1738  gio = 11808  vio = 55192  value = 33552  Temperature = 25.12612
8  Ratio = 0.995217  rln = -0.004794                                            
duration = 1986  gio = 11820  vio = 55205  value = 33547  Temperature = 25.08368
0  Ratio = 0.996824  rln = -0.003181                                            
duration = 1994  gio = 11818  vio = 55213  value = 33550  Temperature = 25.08368
0  Ratio = 0.996825  rln = -0.003180                                            
duration = 2122  gio = 11813  vio = 55197  value = 33553  Temperature = 25.11645
6  Ratio = 0.995584  rln = -0.004426                                            
duration = 2258  gio = 11814  vio = 55219  value = 33570  Temperature = 25.12973
1  Ratio = 0.995082  rln = -0.004930                                            
duration = 2386  gio = 11812  vio = 55210  value = 33564  Temperature = 25.12854
1  Ratio = 0.995127  rln = -0.004885                                            
duration = 2514  gio = 11805  vio = 55196  value = 33550  Temperature = 25.12005
6  Ratio = 0.995447  rln = -0.004563                                            

evanh

It's a rather strong argument to use a timer to sample after a millisecond instead of iterating on decimations.

evanh

@chintan_joshi said:
Still its giving same output with thermistor.

Time to use a multimeter to measure the voltage at the pin.

chintan_joshi

@evanh said:

@chintan_joshi said:
Still its giving same output with thermistor.

Time to use a multimeter to measure the voltage at the pin.

Yes i can able to see voltage starts from 2.5 v, and when i increase room temperature voltage dropping to 1.9v, but still P2 showing 25c

evanh

Oh, maybe because pin P14 is package pin 24?

EDIT: I'm definitely getting voltage swing on the assigned pin and a reading change to match. Eg: with bias at 2.926 V

      pin P8
duration = 1425  gio = 12834  vio = 55901  value = 52585  Temperature = 108.654390  Ratio = 0.083419  rln = -2.483875
duration = 1538  gio = 12842  vio = 55892  value = 53197  Temperature = 118.495240  Ratio = 0.066782  rln = -2.706316
duration = 1642  gio = 12835  vio = 55928  value = 52476  Temperature = 106.810370  Ratio = 0.087082  rln = -2.440910
duration = 1778  gio = 12842  vio = 55914  value = 52066  Temperature = 101.787330  Ratio = 0.098103  rln = -2.321736
duration = 1906  gio = 12829  vio = 55919  value = 51769  Temperature = 98.374050  Ratio = 0.106574  rln = -2.238914
duration = 2058  gio = 12856  vio = 55898  value = 51610  Temperature = 96.848910  Ratio = 0.110647  rln = -2.201414
duration = 2210  gio = 12840  vio = 55909  value = 51502  Temperature = 95.647860  Ratio = 0.113988  rln = -2.171663
duration = 2338  gio = 12836  vio = 55905  value = 51414  Temperature = 94.802490  Ratio = 0.116413  rln = -2.150607
duration = 2442  gio = 12833  vio = 55905  value = 51393  Temperature = 94.597020  Ratio = 0.117013  rln = -2.145474
duration = 2586  gio = 12814  vio = 55919  value = 51360  Temperature = 94.168400  Ratio = 0.118274  rln = -2.134749
duration = 2706  gio = 12832  vio = 55907  value = 51321  Temperature = 93.873840  Ratio = 0.119151  rln = -2.127365
duration = 2842  gio = 12825  vio = 55923  value = 51319  Temperature = 93.722960  Ratio = 0.119603  rln = -2.123577
duration = 2954  gio = 12835  vio = 55896  value = 51301  Temperature = 93.771920  Ratio = 0.119456  rln = -2.124806
duration = 3098  gio = 12845  vio = 55916  value = 51333  Temperature = 93.898870  Ratio = 0.119076  rln = -2.127992
duration = 3218  gio = 12847  vio = 55905  value = 51286  Temperature = 93.536590  Ratio = 0.120164  rln = -2.118895
duration = 3354  gio = 12823  vio = 55905  value = 51301  Temperature = 93.706420  Ratio = 0.119653  rln = -2.123162
duration = 3490  gio = 12827  vio = 55904  value = 51297  Temperature = 93.672180  Ratio = 0.119756  rln = -2.122302

evanh

Ha, need to invert to match your voltages ... bias now at 0.352 V with inverted calculation:

      pin P8
duration = 1425  gio = 12838  vio = 55880  value = 14306  Temperature = 149.719940  Ratio = 0.035311  rln = -3.343575
duration = 1602  gio = 12809  vio = 55879  value = 15536  Temperature = 117.946900  Ratio = 0.067595  rln = -2.694216
duration = 1642  gio = 12804  vio = 55870  value = 16255  Temperature = 106.794640  Ratio = 0.087114  rln = -2.440543
duration = 1770  gio = 12796  vio = 55870  value = 16674  Temperature = 101.434930  Ratio = 0.098939  rln = -2.313255
duration = 1906  gio = 12818  vio = 55876  value = 16939  Temperature = 98.657500  Ratio = 0.105838  rln = -2.245850
duration = 2074  gio = 12834  vio = 55893  value = 17125  Temperature = 96.835200  Ratio = 0.110684  rln = -2.201076
duration = 2194  gio = 12804  vio = 55875  value = 17230  Temperature = 95.457120  Ratio = 0.114530  rln = -2.166921
duration = 2330  gio = 12818  vio = 55872  value = 17276  Temperature = 95.116880  Ratio = 0.115504  rln = -2.158449
duration = 2450  gio = 12794  vio = 55869  value = 17324  Temperature = 94.422250  Ratio = 0.117525  rln = -2.141104
duration = 2586  gio = 12798  vio = 55889  value = 17380  Temperature = 93.929290  Ratio = 0.118985  rln = -2.128756
duration = 2698  gio = 12794  vio = 55891  value = 17380  Temperature = 93.896610  Ratio = 0.119083  rln = -2.127936
duration = 2842  gio = 12817  vio = 55875  value = 17390  Temperature = 93.982850  Ratio = 0.118826  rln = -2.130099
duration = 2978  gio = 12802  vio = 55880  value = 17389  Temperature = 93.867220  Ratio = 0.119171  rln = -2.127199
duration = 3106  gio = 12816  vio = 55875  value = 17402  Temperature = 93.857300  Ratio = 0.119200  rln = -2.126949
duration = 3226  gio = 12812  vio = 55866  value = 17416  Temperature = 93.677430  Ratio = 0.119740  rln = -2.122433
duration = 3354  gio = 12803  vio = 55882  value = 17413  Temperature = 93.645230  Ratio = 0.119837  rln = -2.121624
duration = 3482  gio = 12814  vio = 55875  value = 17418  Temperature = 93.684690  Ratio = 0.119718  rln = -2.122616

chintan_joshi

@evanh said:
Ha, need to invert to match your voltages ... bias now at 0.352 V with inverted calculation:

      pin P8
duration = 1425  gio = 12838  vio = 55880  value = 14306  Temperature = 149.719940  Ratio = 0.035311  rln = -3.343575
duration = 1602  gio = 12809  vio = 55879  value = 15536  Temperature = 117.946900  Ratio = 0.067595  rln = -2.694216
duration = 1642  gio = 12804  vio = 55870  value = 16255  Temperature = 106.794640  Ratio = 0.087114  rln = -2.440543
duration = 1770  gio = 12796  vio = 55870  value = 16674  Temperature = 101.434930  Ratio = 0.098939  rln = -2.313255
duration = 1906  gio = 12818  vio = 55876  value = 16939  Temperature = 98.657500  Ratio = 0.105838  rln = -2.245850
duration = 2074  gio = 12834  vio = 55893  value = 17125  Temperature = 96.835200  Ratio = 0.110684  rln = -2.201076
duration = 2194  gio = 12804  vio = 55875  value = 17230  Temperature = 95.457120  Ratio = 0.114530  rln = -2.166921
duration = 2330  gio = 12818  vio = 55872  value = 17276  Temperature = 95.116880  Ratio = 0.115504  rln = -2.158449
duration = 2450  gio = 12794  vio = 55869  value = 17324  Temperature = 94.422250  Ratio = 0.117525  rln = -2.141104
duration = 2586  gio = 12798  vio = 55889  value = 17380  Temperature = 93.929290  Ratio = 0.118985  rln = -2.128756
duration = 2698  gio = 12794  vio = 55891  value = 17380  Temperature = 93.896610  Ratio = 0.119083  rln = -2.127936
duration = 2842  gio = 12817  vio = 55875  value = 17390  Temperature = 93.982850  Ratio = 0.118826  rln = -2.130099
duration = 2978  gio = 12802  vio = 55880  value = 17389  Temperature = 93.867220  Ratio = 0.119171  rln = -2.127199
duration = 3106  gio = 12816  vio = 55875  value = 17402  Temperature = 93.857300  Ratio = 0.119200  rln = -2.126949
duration = 3226  gio = 12812  vio = 55866  value = 17416  Temperature = 93.677430  Ratio = 0.119740  rln = -2.122433
duration = 3354  gio = 12803  vio = 55882  value = 17413  Temperature = 93.645230  Ratio = 0.119837  rln = -2.121624
duration = 3482  gio = 12814  vio = 55875  value = 17418  Temperature = 93.684690  Ratio = 0.119718  rln = -2.122616

Oh god, issue was the pin i was using. If i changed pin from 14 to 55 then its showing proper temperature of cartridge thermistor as well of room temp thermistor. Thank you very much for your help.

Even if i use below code i am getting proper temperature. Also able to get proper temperature with your code as well.

void tsr_thread()
{
        while(true)
        {
        float samples[5];
        unsigned int beta = 3380;
        for (int i=0; i< 5; i++) 
        {
        samples[i] = (uint32_t)_rdpin(Temp_pin);
        _waitms(300);
    }

      double average = 0;
    for (int i=0; i< 5; i++) 
    {
            average += samples[i];
            printf("\n%d> %f\n",i,samples[i]);
    }

    average /= 5;
    printf("average = %f\n",average);
    average = ((16383 / average) - 1);
    //printf("average = %f\n",average);
    average = 10000 / average;
    float steinhart;
    steinhart = average / 10000;     // (R/Ro)
    steinhart = log(steinhart);                  // ln(R/Ro)
    steinhart /= beta;                   // 1/B * ln(R/Ro)
    steinhart += 1.0 / (25 + 273.15); // + (1/To)
    steinhart = 1.0 / steinhart;                 // Invert
    steinhart -= 273.15; 
    //if(steinhart<25)
    //enable_substrate();
     printf("Temperature = %f\n",steinhart);
     }

}

evanh

Cool, mystery solved at least.

The pin group P12..P15, and however many pins are on the same VIO regulator, will all be suspect. Presumably they have been over-voltaged at some time. I have two groups of eight failed pins on my older "glob-top" Eval Board.

chintan_joshi

Hello @evanh i am trying to get quadrature Encoder readings in every 2 ms, to adjust the DC motor Duty cycle.

I am able to get quadrature encoder readings by this setup

_pinstart(QUADA, P_PLUS1_B | P_QUADRATURE, 0, 0); // QUADA is pin 42

Now is there any timer interrupt/ Method available for P2 using c++ to get encoder readings on every 2 ms?

i am thinking to implement this as below.

void get_position()
{
        int count_pos = 0;
      while(_pollatn()==0);
      while(true)
      {  
        buf_pos[count_pos] = getPulses();// get Quadrature encoder readings
        count_pos = count_pos + 1;
        _waitms(2);
      }
}

evanh

If you want to write the timer code in C then just poll the smartpin to detect next period. Eric has stated there is no attempt to make the FlexC's compiled code interrupt safe. So no support for hooking an ISR.

The quadrature counter smartpin mode can be configured to generate deltas on a metronomic period. All you do is set X to the period, in sysclock ticks, then IN flags high when the period expires. And reading IN automatically clears the flag each time. So each time you get a high you know the next delta count has been produced by the smartpin.

chintan_joshi

@evanh said:
If you want to write the timer code in C then just poll the smartpin to detect next period. Eric has stated there is no attempt to make the FlexC's compiled code interrupt safe. So no support for hooking an ISR.

The quadrature counter smartpin mode can be configured to generate deltas on a metronomic period. All you do is set X to the period, in sysclock ticks, then IN flags high when the period expires. And reading IN automatically clears the flag each time. So each time you get a high you know the next delta count has been produced by the smartpin.

Yes Thank you, but is there any reference or document or code available for P2 to do PID control ? final outcome we want is to control the Dc motor in closed loop based on encoder readings.

it is similar to this link, https://learn.parallax.com/tutorials/language/pbasic/pid-control/proportional-integral-and-derivative

Only difference is we can use encoder readings and adjust PWM duty cycle of Dc motor.

evanh

ah, that's kind of a whole program thing.

Here's a quick demo of just using the smartpin to time a polling loop (for delta accumulation to 64-bit, and a second smartpin for absolute count):

PS: And the pins are config'd with pull-ups for use with open-collector NPN wired encoder.

evanh

@chintan_joshi said:
Only difference is we can use encoder readings and adjust PWM duty cycle of Dc motor.

Is the PWM direct to the power transistors? Or is it for effective torque? If the latter then PDM (NCO Duty mode) is more responsive option than PWM.

PS: I'm assuming you have an opto-coupler in between and thereby need it pulsed rather than using a DAC.

whicker

Here is a closed-loop DC brushed motor project I made that can be used as a starting point.

https://github.com/Haggarman/Demo_PID_Loop

chintan_joshi

@evanh said:

@chintan_joshi said:
Only difference is we can use encoder readings and adjust PWM duty cycle of Dc motor.

Is the PWM direct to the power transistors? Or is it for effective torque? If the latter then PDM (NCO Duty mode) is more responsive option than PWM.

PS: I'm assuming you have an opto-coupler in between and thereby need it pulsed rather than using a DAC.

No, i am using cytron Driver(https://robu.in/product/dual-channel-enhanced-13amp-dc-motor-driver-30a-peak/) to drive the DC motor which is attached to the printer cartridge. To drive the DC motor with cytron driver i need two pins 1. Direction and 2. PWM , with PWM pin i can set require speed of the motor. So on normal operation i can drive the Cartridge motor with PWM duty cycle. in this scenario there is no real time checking for cartridge position.

Now i need to update speed(PWM duty ) in closed loop to check in real time for cartridge position using encoder, So need to check cartridge position every 2 ms and update Speed to make it equal distance travel for every 2 ms.

For arduino there is one PID library which can continuously check Encoder position and update PWM duty appropriately, so looking for similar solution for P2.

chintan_joshi

@whicker said:
Here is a closed-loop DC brushed motor project I made that can be used as a starting point.

https://github.com/Haggarman/Demo_PID_Loop

Hey @whicker, yes this looks similar solution, but not able to understand code due to spin2 language. is there any c++ alternative for this for P2?

i am using only 1 Dc motor motor and 1 encoder.

#define QUADA 42 // Encoder pin A
#define QUADB 43 // Encoder pin B
#define M1_IN1 28 // Cytron driver Direction pin
#define M1_IN2 29 // Cytron Driver PWM pin

And in Demo_PILoop5.spin2 file different pin_offset and app io pins are declared for 4 motors and 4 encoders

CON { pin offsets }
  HBRIDGE_NFault = 0  { IO }                            'open drain pulls down when fault
  HBRIDGE_Enable = 1  { O }
  HBRIDGE_AIN_1 =  2  { O }
  HBRIDGE_AIN_2 =  3  { O }
  HBRIDGE_BIN_2 =  4  { O }
  HBRIDGE_BIN_1 =  5  { O }

  RIGHT_MOTORS =   0
  LEFT_MOTORS =   16

CON { app io pins }
  LEFT_MOTORS_FAULT = LEFT_MOTORS + HBRIDGE_NFault
  LEFT_MOTORS_ENABLE = LEFT_MOTORS + HBRIDGE_Enable

  LEFT_REAR_FWD = LEFT_MOTORS + HBRIDGE_AIN_1
  LEFT_REAR_REV = LEFT_MOTORS + HBRIDGE_AIN_2
  LEFT_REAR_ENCA = 8
  LEFT_REAR_ENCB = 9


  LEFT_FRONT_FWD = LEFT_MOTORS + HBRIDGE_BIN_1
  LEFT_FRONT_REV = LEFT_MOTORS + HBRIDGE_BIN_2
  LEFT_FRONT_ENCA = 10
  LEFT_FRONT_ENCB = 11


  RIGHT_MOTORS_FAULT = RIGHT_MOTORS + HBRIDGE_NFault
  RIGHT_MOTORS_ENABLE = RIGHT_MOTORS + HBRIDGE_Enable

  RIGHT_REAR_FWD = RIGHT_MOTORS + HBRIDGE_AIN_2
  RIGHT_REAR_REV = RIGHT_MOTORS + HBRIDGE_AIN_1
  RIGHT_REAR_ENCB = 12
  RIGHT_REAR_ENCA = 13

  RIGHT_FRONT_FWD = RIGHT_MOTORS + HBRIDGE_BIN_2
  RIGHT_FRONT_REV = RIGHT_MOTORS + HBRIDGE_BIN_1
  RIGHT_FRONT_ENCB = 14
  RIGHT_FRONT_ENCA = 15

To use this file i have to call MAIN() of Demo_PILoop.spin2 from c++ code, but before that i have to change pins for 1 motor in spin2 code. and not able to understand What this pin offset and app io pins are declared for?

chintan_joshi

@whicker , i am trying to update code for single motor and encoder.

i have made below changes , but not able to understand where is motor DIR and PWM pins declared in spin2 code, where can i modify motor DIR and PWM pins to 28 and 29? I have changed Encoder pins in code and changed AXIS_COUNT to 1 and commented out other 3 motors code . can you please help to make this code for single motor and encoder?

#define M1_IN1 28 // Cytron driver Direction pin
#define M1_IN2 29 // Cytron Driver PWM pin

''01/17/2022
''rewired to use adafruit DRV8833
''for what it's worth, it's now much less sample time dependent.
''mess with PILOOP_SAMPLES_Hz and speed setpoint
''speed command target_speed_f is now in [motor revolutions per second]
''re-enabled tracking time constant after bug discovered by setting a negative literal value in a pub section.
''as workaround, use the constant "TARGET_SPEED_FLOAT" to change desired motor speed.
''ramp up and down. encoder has more smoothing (at the cost of lag)
''create a PILoop object
''run all 4 motors at the same time

CON { timing }

  CLK_FREQ = 200_000_000                                ' system freq as a constant
  BR_TERM  = 230_400                                    ' terminal baud rate

  _clkfreq = CLK_FREQ                                   ' set system clock

CON { fixed io pins }

  RX1      = 63  { I }                                  ' programming / debug
  TX1      = 62  { O }

  SF_CS    = 61  { O }                                  ' serial flash
  SF_SCK   = 60  { O }
  SF_SDO   = 59  { O }
  SF_SDI   = 58  { I }

  SD_SCK   = 61  { O }                                  ' sd card
  SD_CS    = 60  { O }
  SD_SDI   = 59  { O }
  SD_SDO   = 58  { I }

  'SDA1     = 57  { IO }                                ' i2c (optional)
  'SCL1     = 56  { IO }

  LED_user2 = 57
  LED_user1 = 56

CON { pin offsets }
  HBRIDGE_NFault = 0  { IO }                            'open drain pulls down when fault
  HBRIDGE_Enable = 1  { O }
  HBRIDGE_AIN_1 =  2  { O }
  HBRIDGE_AIN_2 =  3  { O }
  HBRIDGE_BIN_2 =  4  { O }
  HBRIDGE_BIN_1 =  5  { O }

  RIGHT_MOTORS =   0
  LEFT_MOTORS =   16

CON { app io pins }
  LEFT_MOTORS_FAULT = LEFT_MOTORS + HBRIDGE_NFault
  LEFT_MOTORS_ENABLE = LEFT_MOTORS + HBRIDGE_Enable

  LEFT_REAR_FWD = LEFT_MOTORS + HBRIDGE_AIN_1
  LEFT_REAR_REV = LEFT_MOTORS + HBRIDGE_AIN_2
  LEFT_REAR_ENCA = 42
  LEFT_REAR_ENCB = 43


  ''LEFT_FRONT_FWD = LEFT_MOTORS + HBRIDGE_BIN_1
  ''LEFT_FRONT_REV = LEFT_MOTORS + HBRIDGE_BIN_2
  ''LEFT_FRONT_ENCA = 10
  ''LEFT_FRONT_ENCB = 11


  ''RIGHT_MOTORS_FAULT = RIGHT_MOTORS + HBRIDGE_NFault
  ''RIGHT_MOTORS_ENABLE = RIGHT_MOTORS + HBRIDGE_Enable

  ''RIGHT_REAR_FWD = RIGHT_MOTORS + HBRIDGE_AIN_2
  ''RIGHT_REAR_REV = RIGHT_MOTORS + HBRIDGE_AIN_1
  ''RIGHT_REAR_ENCB = 12
  ''RIGHT_REAR_ENCA = 13

  ''RIGHT_FRONT_FWD = RIGHT_MOTORS + HBRIDGE_BIN_2
  ''RIGHT_FRONT_REV = RIGHT_MOTORS + HBRIDGE_BIN_1
  ''RIGHT_FRONT_ENCB = 14
  ''RIGHT_FRONT_ENCA = 15


CON { user }
  AXIS_COUNT = 1               'number of motors
  MOTORS_PWM_Hz = 5000          'the pwm motor frequency

  PILOOP_SAMPLES_Hz = 20        'TRY 5, 10, 20, 50. it should be better behaved.
  PILOOP_WAIT_COUNT = CLK_FREQ / PILOOP_SAMPLES_Hz

  SPEED_AT_FULL_PWM = 160        'motor conversion factor speed command to PWM (Make it a speed reachable at 100% PWM in practice. Reduce from measured freewheel speed somewhat to include some expected load.)

  K_Val = 2.5     'Increasing this Proportional value pushes back harder and harder against error. 1.0 is unity, so one unit of error produces one unit of pushback.
  P_Weight = 0.55 'Optional Weighting value, leave at 1.0 for normal PID overshoot type behavior. A value like 0.5 significantly reduces initial overshoot, but only after correct tuning of K and I.
  I_Val = 2.0     'Increasing this value makes the Integration effect stronger, but more likely to cause oscillation. Too small and you never reach setpoint. This is technically 1/I from textbook definition.
  TrackTC = 2.0   'Anti-windup filter "time constant" to saturate Integrator . It is the same time unit as your process value (seconds). Higher values allow more Integrator windup (which is bad).

CON {settings}
  TARGET_SPEED_FLOAT = 100.0      'motor [rev/sec] SPEED SETPOINT. Ranges from about -20.0 to -160.0 and 20.0 to 160.0 (static friction break requires some deadzone, it'll be start-stop below 20.0).

OBJ
  MOT[AXIS_COUNT] : "Motor_DRV8833.spin2"
  ENC[AXIS_COUNT] : "Motor_ABEncoder3.spin2"
  PID[AXIS_COUNT] : "Motor_PILoop.spin2"

PUB MAIN() | waitForCnt, add_f, target_speed_f, actual_speed_f[4], command_pwm[4]
  pinlow(LED_user1)

  ''Initialization of the motors and encoders:
  pullup(LEFT_MOTORS_FAULT, P_HIGH_15K)
  ''pullup(RIGHT_MOTORS_FAULT, P_HIGH_15K)
  waitms(100)
  pinhigh(LED_user1)
  MOT[0].start(LEFT_REAR_FWD, LEFT_REAR_REV, MOTORS_PWM_Hz)
  ''MOT[1].start(LEFT_FRONT_FWD, LEFT_FRONT_REV, MOTORS_PWM_Hz)
  ''MOT[2].start(RIGHT_REAR_FWD, RIGHT_REAR_REV, MOTORS_PWM_Hz)
  ''MOT[3].start(RIGHT_FRONT_FWD, RIGHT_FRONT_REV, MOTORS_PWM_Hz)

  ENC[0].start(LEFT_REAR_ENCA, LEFT_REAR_ENCB, 4, CLK_FREQ)
  ''ENC[1].start(LEFT_FRONT_ENCA, LEFT_FRONT_ENCB, 4, CLK_FREQ)
  ''ENC[2].start(RIGHT_REAR_ENCA, RIGHT_REAR_ENCB, 4, CLK_FREQ)
  ''ENC[3].start(RIGHT_FRONT_ENCA, RIGHT_FRONT_ENCB, 4, CLK_FREQ)

  PID[0].start(PILOOP_SAMPLES_Hz, MOT.PWM_REF_MAX, SPEED_AT_FULL_PWM)
  ''PID[1].start(PILOOP_SAMPLES_Hz, MOT.PWM_REF_MAX, SPEED_AT_FULL_PWM)
  ''PID[2].start(PILOOP_SAMPLES_Hz, MOT.PWM_REF_MAX, SPEED_AT_FULL_PWM)
  ''PID[3].start(PILOOP_SAMPLES_Hz, MOT.PWM_REF_MAX, SPEED_AT_FULL_PWM)

  waitms(100)
  pinlow(LED_user1)
  pinhigh(LEFT_MOTORS_ENABLE)
  ''pinhigh(RIGHT_MOTORS_ENABLE)


  ''Some commented out code here for open-loop speed measurement:
  'target_pwm := -MOT1.PWM_REF_MAX / 2                    '50 percent
  'MOT1.setpwm(target_pwm)
  'waitms(100)
  'openloop()

  'debug(`SCOPE MyScope SIZE 800 600 SAMPLES 128)
  'debug(`MyScope 'PWM' 0 10000 500 10 %1111)
  'debug(`MyScope 'Speed' 0 160 500 20 %1111)
  'debug(`MyScope 'Setpoint' 0 160 500 20 %1111)

  ''PI Controller Loop Initialization:
  waitForCnt := getct() + PILOOP_WAIT_COUNT
  actual_speed_f[0] := ENC[0].update()
  ''actual_speed_f[1] := ENC[1].update()
  ''actual_speed_f[2] := ENC[2].update()
  ''actual_speed_f[3] := ENC[3].update()

  target_speed_f := TARGET_SPEED_FLOAT
  add_f := 0.75
  repeat
    ''Ramp speed up and down (just as a test)
    target_speed_f := target_speed_f +. add_f
    if target_speed_f >. 150.0
      add_f := 0.0 -. add_f
    elseif target_speed_f <. -.150.0
      add_f := 0.0 -. add_f

    ''wait until next loop update
    waitct(waitForCnt += PILOOP_WAIT_COUNT)
    actual_speed_f[0] := ENC[0].update()
    ''actual_speed_f[1] := ENC[1].update()
    ''actual_speed_f[2] := ENC[2].update()
    ''actual_speed_f[3] := ENC[3].update()

    command_pwm[0] := PID[0].update(target_speed_f, actual_speed_f[0] )
    ''command_pwm[1] := PID[1].update(target_speed_f, actual_speed_f[1] )
    ''command_pwm[2] := PID[2].update(target_speed_f, actual_speed_f[2] )
    ''command_pwm[3] := PID[3].update(target_speed_f, actual_speed_f[3] )

    MOT[0].setpwm(command_pwm[0] )
    ''MOT[1].setpwm(command_pwm[1] )
    ''MOT[2].setpwm(command_pwm[2] )
    ''MOT[3].setpwm(command_pwm[3] )


PUB openloop(MY_AXIS) | waitForCnt, actual_speed_f, debug1
''this just reports the current speed in a loop
  waitForCnt := getct() + PILOOP_WAIT_COUNT
  repeat
    actual_speed_f := ENC[MY_AXIS].update()
    debug1 := round(actual_speed_f)
    debug(sdec(debug1))
    waitct(waitForCnt += PILOOP_WAIT_COUNT)


PUB pullup(pinfield, type)
''pinfield = just a single pin number, or the field of pins which is a starting pin and the quantity of pins after it.
''type = [P_HIGH_FAST, P_HIGH_1K5, P_HIGH_15K, P_HIGH_150K, P_HIGH_1MA, P_HIGH_100UA, P_HIGH_10UA, P_HIGH_FLOAT]
  wrpin(pinfield, type)
  pinhigh(pinfield)

evanh

@chintan_joshi said:

@evanh said:
Is the PWM direct to the power transistors? Or is it for effective torque? If the latter then PDM (NCO Duty mode) is more responsive option than PWM.

No, i am using cytron Driver(https://robu.in/product/dual-channel-enhanced-13amp-dc-motor-driver-30a-peak/) to drive the DC motor which is attached to the printer cartridge. To drive the DC motor with cytron driver i need two pins 1. Direction and 2. PWM , with PWM pin i can set require speed of the motor.

That's option one, the PWM directly modulates the motor.
It should have optocouplers but doesn't.

evanh

@chintan_joshi said:
For arduino there is one PID library which can continuously check Encoder position and update PWM duty appropriately, so looking for similar solution for P2.

We can probably use a lot of it. Is there a link?

chintan_joshi

@evanh said:

@chintan_joshi said:
For arduino there is one PID library which can continuously check Encoder position and update PWM duty appropriately, so looking for similar solution for P2.

We can probably use a lot of it. Is there a link?

This is closed match for PID control what @whicker suggested. But its in spin2. Let me check for Arduino Link.

chintan_joshi

@chintan_joshi said:

@evanh said:

@chintan_joshi said:
For arduino there is one PID library which can continuously check Encoder position and update PWM duty appropriately, so looking for similar solution for P2.

We can probably use a lot of it. Is there a link?

This is closed match for PID control what @whicker suggested. But its in spin2. Let me check for Arduino Link.

Found the Arduino PID library link, https://github.com/br3ttb/Arduino-PID-Library

https://projecthub.arduino.cc/tareqwaleed1996/c9f9b99c-fe07-4250-8799-b81fbf9260da

evanh

I just looked at Whicker's code. It is a velocity control loop rather than positional. I suspect you'll be wanting positional control loop.

chintan_joshi

@evanh said:
I just looked at Whicker's code. It is a velocity control loop rather than positional. I suspect you'll be wanting positional control loop.

Yes but he is controlling velocity based on encoder readings, so i thought its same. Because if we can run motor on constant velocity based on encoder readings, then it will do the job to run the cartridge at constant speed.

evanh

@chintan_joshi said:
... then it will do the job to run the cartridge at constant speed.

You'll be wanting to know the location of the printhead too I suspect. The usual way is to have a profile generator ramp up from an exact start position, reaching traverse speed, then do print, possibly also at exact locations, then ramp back down to an exact end position.

So the PID function is just a small part of the servo.

evanh

The attached C isn't useable by itself but compiles without error and should work when the rest is done. I've converted all the .h C++ privates into equivalent .c statics and replaced the millisecond function call with same flexC function: _getms().