0

I have recently written a program which is supposed to change the duty-cycle to keep the feedback voltage measured on A0, A1 constant. The problem is that the Arduino will start crashing at pretty constant intervals if it approaches it's target current. Once it crashes I have confirmed that it restarts the program (but crashes again in a moment).

Here's a sample of the serial output during the crash:

Current A: 65   Current B: 67   PWM A: 174.98   PWM B: 236.00
Current A: 65   Current B: 67   PWM A: 174.98   PWM B: 235.96
Current A: 65   Current B: 67   PWM A: 174.98   PWM B: 235.92
Current A: 64   Current B: 67   PWM A: 175.00   PWM B: 235.88
Current A: 64   Current B: 66   PWM A: 250.00   PWM B: 285114740000000000000000000.00
Current A: 0    Current B: 0    PWM A: 1.30     PWM B: 0.00
Current A: 0    Current B: 0    PWM A: 2.60     PWM B: 1.30
Current A: 0    Current B: 0    PWM A: 3.90     PWM B: 2.60
Current A: 0    Current B: 0    PWM A: 5.20     PWM B: 3.90

And here is all of my code:

#include <MemoryFree.h>

//Program to run on the Vegetarium
//Written by Leon Teichroeb on 12/6/2016

//Hardware description:
//-Oscillator 2A and 2B for the two onboard driver circuits
//-Pin 13 is attached to the "heartbeat" LED
//-Channel A (3) and B (11) current measurements on A0 and A1

//Functionality:
//-Drive Channels at set current

float PWMA = 0;
float PWMB = 0;

int TARGETA = 65;
int TARGETB = 65;


void setup() {
  Serial.begin(115200);

  pinMode(3, OUTPUT);
  pinMode(11, OUTPUT);
  TCCR2A = _BV(COM2A1) | _BV(COM2B1) | _BV(WGM21) | _BV(WGM20);
  TCCR2B = _BV(CS20);
  OCR2A = (int)PWMA;
  OCR2B = (int)PWMB;
  digitalWrite(3, LOW);
  digitalWrite(11, LOW);

  delay(5000);
}

void loop() {
  int CURRENTA = analogRead(A0) / 4;

  if(CURRENTA < TARGETA) {
    PWMA += (TARGETA - CURRENTA) * 0.020f;
    if(PWMA > 250) {
      PWMA = 250;
    }
  }
  if(CURRENTA > TARGETA) {
    PWMA += (TARGETA - CURRENTA) * 0.020f;
    if(PWMA < 5) {
      PWMA = 5;
    }
  }

  int CURRENTB = analogRead(A1) / 4;

  Serial.println("Current A: " + (String)CURRENTA + "\t" + "Current B: " + (String)CURRENTB + "\t" + "PWM A: " + (String)PWMA + "\t" + "PWM B: " + (String)PWMB);
  //Serial.println("PWM A: " + (String)PWMA + "\t" + "PWM B: " + (String)PWMB);
  //Serial.print("freeMemory()=");
  //Serial.println(freeMemory());


  if(CURRENTB < TARGETB) {
    PWMB += (TARGETB - CURRENTB) * 0.020f;
    if(PWMB > 250) {
      PWMB = 250;
    }
  }
  if(CURRENTB > TARGETB) {
    PWMB += (TARGETB - CURRENTB) * 0.020f;
    if(PWMB < 5) {
      PWMB = 5;
    }
  }

  if(PWMA < 10) {
    digitalWrite(3, LOW);
  } else {
    analogWrite(3, (int)PWMA);
  }
  if(PWMB < 10) {
    digitalWrite(11, LOW);
  } else {
    analogWrite(11, (int)PWMB);
  }

  delay(40);
}

My circuit also provides power for the Arduino through a simple 7815 voltage regulator. The analog input does not exceed a few volts, and the PWM outputs are driving two logic level MOSFETs. If more information on the circuit is necessary please let me know!

EDIT: I have found out that it isn't a bug in the two feedback loops, because both work on their own. It crashes as soon as I use both feedback loops together. As far as I'm aware, there are no shared variables or data between them

5
  • 1
    I can't see anything horribly wrong with your code. Maybe a memory issue. Did you try printing without Strings? Like: Serial.print("Current A: "); Serial.print(CURRENTA); .... It's cumbersome but memory friendly. Jun 18, 2016 at 11:56
  • I can definitely give it a try, but I actually wrote the serial code to debug this problem specifically. I also suspected memory issues, but with MemoryFree.h I found that I still have >1500 bytes free and no leaks
    – lte678
    Jun 18, 2016 at 13:09
  • The 285114740000000000000000000 is kind of odd. Also, PWM B: 0.00 seems to be impossible with the code above. Very odd. (Did you add some capacitors to the output of the 7805?)
    – Gerben
    Jun 18, 2016 at 13:55
  • The 0.0 PWM signal is possible because it is only limited to 5.0 if it is lowering the duty-cycle. Also, I have a 0.1uF capacitor on the output of the 7805 even though it's powered with a lab bench power supply
    – lte678
    Jun 18, 2016 at 14:03
  • Target current or target voltage?
    – Duncan C
    Apr 6, 2019 at 12:10

1 Answer 1

1

I finally figured it out, the Arduino was entering a current protection mode causing it to reset. The pins driving the MOSFETs were pulling too much current once the regulation started because of the BUZ11's input capacitance.

To fix the problem I just had to add 120 ohm resistors in series with the gate. This made the activation and deactation slower resulting in less current flow.

2
  • are u telling u have added 120 ohm resistor in series with ic 7815 i am having same prob. please tell is this sol. will work Apr 6, 2019 at 11:26
  • What's a BUZ11, a MOSFET? So the gate input capacitance on the MOSFET was exceeding the current limits on the logic lines? Or are you talking about the power draw of the MOSFET? Can you provide a wiring diagram? It's hard to understand what you're saying.
    – Duncan C
    Apr 6, 2019 at 12:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.