Electronic – Why do all the LEDs in this circuit light up at once

ledpwm

I have two things to say before I ask my question:

I am a layman with very little theoretical understanding of electronics (my question text will prove this, trust me). Sorry if my idea of the problem is a weird mess.
Coming from StackOverflow, I know that asking a question about such a specific scenario/problem may not add much value for others. I really just need help understanding what's going on here and I hope there is some general understanding in the end that might also help others.

What I want to achieve

I want to build a little box, exposing three rotary dials (or potentiometers in my case), each controlling the brightness of one channel of a single RGB-LED (common ground).

What I have tried

I bought three simple PWM-Modules (link to the product page for reference, not a referrer link!), each providing a potentiometer and +/- connections for the DC in and "motor" out (they're primarily meant for motors). I don't have the schematics of those, so my sketch below shows an abstraction of the modules. I also bought a little RGB-LED module that's nothing more than the LED and three resistors (pretty superfluous).

Here is a plan of how I prototyped my build:

The problem

For me (as an absolute layman) this looks like it should work. But the problem is that as soon as I turn up one of the PWM pots to control a single LED (of the RGB LED package), all three LEDs light up (nice and slow, so at least that works).

When I disconnect two of the LEDs, only one works (obviously). When I measure the voltage on the disconnected "out +" (from PWM out + to the LED) – all three get the same voltage. How is this possible? I want to control the RGB colors independently.

Is it possible that the PWM modules are not made to work in such a parallel configuration? How can I solve this? Thank you in advance!

PS: I know so little I don't even know how to correctly tag this question 🙁

Best Answer

Your LED array is common-cathode so you cannot use it with these PWM modules. This is inherent to the LED device itself and cannot be changed.

The PWM modules have the + in and + out more-or-less connected together (there's a polyfuse in each line) and that won't work for you.

You can use individual LEDs with resistors. It's also possible to use a common-anode RGB LED + resistors however the fuses will be paralleled, which is not ideal.

The module schematic (simplified) looks like this:

schematic

^{simulate this circuit – Schematic created using CircuitLab}

LED array:

schematic

^{simulate this circuit}

Related Solutions

Multiplexing AVR (ATMega328) PWM within Interrupts

Ok, I figured it out. The problem was that I was leaving the PWM frequency at the default and using a very high multiplexing rate.

By default, the prescalers are set to 64 making the frequency on the timers I was using (0 and 2) 976.5625Hz. But I was using a multiplexing rate of 6400 Hz... just because this was what I used on a different project that didn't include PWM. The main problem here was that it couldn't get any PWM cycles in during a single multiplex cycle.

So, I switched to using a prescale of 8 on my PWMs, bringing the frequency to ~8Khz and set the multiplex rate to 1000Hz, so that way, it gets in about 8 PWM cycles per multiplex cycle. I could probably decrease the multiplex rate even more, but it's a fine line before you start to see flicker.

Also, I found out that when you update the PWM duty cycle (like by setting OCR2A) it take some time before it actually takes effect. So if I move on with the multiplex right away, the PWM values will sort of bleed over into the next LED. So I added some logic where on the first time through the multiplex ISR, it disables all anodes and sets the PWM values. Then on the next time through, it just enables the needed anode. This is what the flip_flop variable is for.

void setup()
{
  //Setup outputs for anodes
  DDRB |= (_BV(PINB0) | _BV(PINB1) | _BV(PINB2));


  //Setup Timer2 Interrupts
  TCCR2A = _BV(COM2A1) | _BV(WGM21) | _BV(WGM20); //Fast PWM mode on OCR2A (PINB3 - Digital 11)
  TCCR2B = _BV(CS21); //8 Prescaler
  OCR2A = 255;

  //Setup Multiplex ISR on Timer1
  // set compare match register for 1000hz 
  OCR1A = 16000;
  // turn on CTC mode
  TCCR1B |= _BV(WGM12);

  TCCR1B |= PRESCALE1_1;  
  // enable timer compare interrupt
  TIMSK1 |= _BV(OCIE1A);
}

volatile uint8_t col = 0;
bool flip_flop = false;
ISR(TIMER1_COMPA_vect)
{
  //Set all anodes to LOW
  if(!flip_flop)
  {
    PORTB &= ~(_BV(PINB0) | _BV(PINB1) | _BV(PINB2));

    //Faking it here: Basically simulating that LED 0 is on but LEDs 1 and 2 are not.
    //Note that because the PWM is on the cathode, 255 is OFF and 0 is full brightness.
    if(col == 0)
      OCR2A = 0;
    else
      OCR2A = 255;
  }



  if(flip_flop)
  {
    //Turn on JUST the LED that we want.
    PORTB |= _BV(col);

    //Step through the LEDs
    col++;
    if(col > 2)
    {
      col = 0;
    } 
  }

  flip_flop = flip_flop ? false : true;
}

Electronic – arduino – RGB LED strip, powered by an ATX PSU and controlled by PWM, makes the speakers buzz with Arduino connected by USB

Please use an opto-isolated device to trigger or replace the FETs and share your results.

This will isolate the (ATX1 + FET + LED) circuit from the (ATX2 + Arduino + speakers) circuit.

Edit:
The power supplies are SMPS operating at different frequencies. You have tied their grounds together and are grounding signals from one through the other. The speakers' V- (ATX1 + ATX2) is fluctuating at a different frequency than its V+ (ATX2 only) so, you hear noise. You may have overlooked that the monitor is grounded through it's own power cable and ATX2 (via the video cable).