Read Vehicle speed sensor

Other Hardware Sensors
1

I wrote program to read number of pulses of Vehicle speed sensor(VSS). It works correctly and read this 5V pulses, but I have problem when there occure some unstability in system and on the pin occure some 0.4V pulse. It count this like a VSS pulse but I would not like read this. Have you an idea how can I filter this unstability?

2

What device is that takes 0.4V as logical 1 level?

3

I think what Blimpy is trying to say is "If it's only 0.4V, why is it messing up your count?"

Try a resistor and capacitor. Google for "RC Network". Do you have an oscilloscope? It will be easier to see what's happening if you do.

4

That 0.4V is not logical 1. It is unbalance on voltage in the car. It occur when I press brake or turn on winkers.

I would like count only pulses from VSS sensor which are 5V, but interrupt count every voltage change on input.
I have only oscilloscope via sound card in PC. I see there that the VSS pulse has 5V and this mess about 0.2 - 0.4V.
I think that RC network don´t help me, because it filter only frequencies and frequency of this mess is variable.
I am affraid if I use RC Network, that system not react on fast VSS pulses. VSS pulse come every 1 feet. It might be one pulse every 0.01 second.
I need to filter signals lower than 0.8 V.

5

Put a print screen of that oscilloscope signal... The RC filter you calculate to pass your pulses frequency and reject higher frequency noise. If any.
It could be a power supply problem.

Schematics and code please. They-re important here. When people see them, they become more able and willing to help.

6

Right signal

False signal

Code

void setup(){
  
    
  noInterrupts(); 
  TCCR1A = 0; 
  TCCR1B |= ((1 << CS10) | (1 << CS11)); 
  TIMSK1 |= (1 << TOIE1); 
  TCNT1 = 3036; 

  pinMode(dist_pin, INPUT);

  attachInterrupt(digitalPinToInterrupt(dist_pin), distance_count, FALLING);
  
  interrupts();
 

}

void loop(){

  
}

ISR(TIMER1_OVF_vect) { 
  counter++; 
  if(counter > 3){

    speed = pulse_distance * dist_pulse_count * 3600;

    if(speed > 3){
      avg_speed_divider++;
      avgSpeed();
      maxSpeed();
    }

    dist_pulse_count = 0;
    counter = 0;
    } 
  TCNT1 = 3036; 
}

void distance_count(){
  dist_pulse_count++;                     // při přerušení přičítá 1 do čítače
  pulse_count++;
  //traveled_distance += pulse_distance;    // při přerušení přičítá ujetou vzdálenost dle délky pulzu
}

void maxSpeed(){
  if(speed > max_speed) max_speed = speed;
}

void avgSpeed(){
  all_speed_values += speed;
  avg_speed = all_speed_values / avg_speed_divider;
}
7

did you tried to read input pin in distance_count() and if it is high just return without incrementing any counter?

eventually with a few microseconds delay, so if it is a valid fall it doesnt read high.

EDIT
or, did you tried to enable interrupt on LOW instead of FALLING?

a low pass filter might smooth out that short glitch, I see your normal pulses are much longer.

8

Dont' use LOW mode for interrupts unless you know what you are doing!!!

Sounds like you are not using star-grounding so will see gournd currents from other circuits. Which
is of course the norm for a vehicle where the chassis is ground.

+0.4V will not register on an Arduino pin as HIGH.

-0.4V on the other hand could be causing bad things to happen.

There may also be very short pulses of a higher voltage happening.

Try adding 1k in series to mitigate this, and 1nF or 10nF to ground on the pin to filter out high speed
glitches. Even a pulse a few 10's of nanoseconds wide can trigger an interrupt, you have
to clean up the incoming signal properly really...

9

Dont' use LOW mode for interrupts unless you know what you are doing!!!

Could you please share a link or short info about these problems? I haven't found examples of LOW interrupt handling and wondered why.

10

Because it interrupts the main program continuously while the input is low. The program never gets a chance to run and the interrupt handler gets 100% of the processor time. Except 50-90% of that is just the interrupt entry/exit action, so you have effectively killed your chip from doing any processing at all while that input stays low.

11

I very much thank all of you. I tried R 1k and C 10 nF and it is good. It gives me good signal. I tried second way. I used diod in series and it is too right function.