Timing Hall Effect Sensor on-off

Projects Programming
1

I'm trying to time a hall effect sensor on-off when a magnet passes. All I get from the output is "0". Tried a number of things but no good results. Sensor is a 44E001. Thanks in advance for the help. Sketch is below:

int oldTime;
const byte interruptPin = 2;
int elapsedTime;

void setup() {
  pinMode(interruptPin, INPUT_PULLUP);
  attachInterrupt(digitalPinToInterrupt(interruptPin), test, RISING);
  Serial.begin(9600);
}

void loop() {
oldTime = millis();

}

void test() {
  elapsedTime = millis() - oldTime;
  Serial.print(" elapsedTime =");
  Serial.print (elapsedTime);
  delay(50);
}
2

millis() won't increment in an interrupt routine. Just set a flag and check in main loop.

...and you shouldn't use delay() in an interrupt either.

1 Like
3

If the suggestions of @red_car do not solve your issues:

Google finds nothing for that number. Post a data sheet for that sensor.

Please post a schematic of the project. Hand drawn, photographed and posted is fine. Include all pin names/numbers, components, their part numbers and/or values and power supplies.

And here are some tutorials on using millis() for timing:
Blink without delay().
Beginner's guide to millis().
Several things at a time.

Don't use Serial.print in an ISR either. Serial.print depends on interrupts and interrupts are disabled in an ISR.

4

there's no need for millis() to increment within the interrupt.
won't the interrupt be able to capture the current value of millis()?

consider (compiles, but not tested)

const byte interruptPin = 2;

byte          cnt;
unsigned long msec;
unsigned long msec0;
unsigned long msecLst;

void test()
{
    msec0 = msec;
    cnt++;
}

void loop()
{
    msec = millis();

    if (cnt)  {
        Serial.println (msec0 - msecLst);
        msecLst = msec0;
        cnt     = 0;
    }
}

void setup() {
    pinMode(interruptPin, INPUT_PULLUP);
    attachInterrupt(digitalPinToInterrupt(interruptPin), test, RISING);
    Serial.begin(9600);
}
5

msec0 and cnt should be declared volatile.

1 Like
6

@PMLAPL
In addition to the other answers - your function will never gives correct elapse time because you do not store anywhere the oldTime value.

7

Why use an interrupt at all? Use StateChangeDetection

const byte hallInputPin = 2;

void setup() 
{
  pinMode(hallInputPin, INPUT_PULLUP);
  Serial.begin(9600);
}

void loop() 
{
  static unsigned long prevMillis = millis();
  static int prevHallState = digitalRead(hallInputPin);
  unsigned long currMillis = millis();
  int currHallState = digitalRead(hallInputPin);
  
  if (currHallState != prevHallState)
  {
    prevHallState = currHallState;
    if (currHallState == HIGH)
    {
      // LOW to HIGH transition.  Start timing.
      prevMillis = millis();
    }
    else
    {
      // HIGH to LOW transition.  Calculate Elapsed Time.
      Serial.print("Elapsed time (ms) = ");
      Serial.println(prevMillis - currMillis);
    }
  }
}
8

I've been trying to learn how to use an interrupt routine since I never used one. I've given up and went back to using state-change that I used to check an IR sensor in a previous sketch. I just switched to a hall effect sensor. Thanks for all the help. Here's my final code that works fine:

unsigned long startTime;
const byte hallInputPin = 5;
unsigned long elapsedTime = 0;
unsigned long Count = 0;
byte Sense;
byte lastSense;

void setup() {
  pinMode(hallInputPin, INPUT);
  startTime = millis();
  Serial.begin(9600); 
}

void loop() {
  Sense = digitalRead(hallInputPin);
  if (lastSense != Sense) {  
    lastSense = Sense;       
    if (Sense == LOW) {      
      Count = Count + 1;
    }
  }
  if (Count == 2) {
    elapsedTime = millis() - startTime;
    startTime = millis();
    Serial.print("Elapsed Time ");
    Serial.print(elapsedTime);
    Serial.println(" milliseconds");
    Count = 0;
  }
}
9

nope - the calculated time has nothing to do with the interval between pulses, because you save the startTime in incorrect moment
You should store the millis() to the startTiime when count =1

10

I think you might be right. There are two magnets on the shaft so then if I check time after three pulses, I should get the time for a full revolution. Thanks for the suggestion.

11

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.