ESP32 + WS2812B wiring

Mike, the above colours, combined with max brightness, were chosen so that red, green and blue channels would be either off or fully on. Do you think that the led current is still pulsed, and that is confusing the OP's multimeter? Could pwm be being used to "limit" the led current, even at full brightness, rather than series resistors or constant current sources?

Goofy3, that green measurement is strange. Mike's theory might explain it but I don't know why the other measurements were not also affected.

The voltage drops you are seeing could be part of the problem. Led strips are not simple Ohmic conductors. Make sure you are connecting them to the power supply with quality, thick wires, not thin wires or dupont wires. Connect both ends of the strip (did you already say that?) and maybe the middle also if you still see a drop at the centre. Some large electrolytic caps might also help.

Ok, thanks guys! In next couple of days I wil connect also middle of the strip. For now only both ends are connected, and I'm using wires that are already soldered to the strip. I have 4700u cap, I will add it to the circuit and we will see what results will be.

PaulRB:
Mike, the above colours, combined with max brightness, were chosen so that red, green and blue channels would be either off or fully on. Do you think that the led current is still pulsed, and that is confusing the OP's

Yes I do think that. So I just did a test. I wired a small strip of WS2812's to the Arduino and turned on only one LED ( green ) at the full 255 value.

Then I put a 10R in the ground and measured across that. This is what I measured:-

Edit. At the time of posting I can't see the image. It flashes on the page and then vanishes. Maybe a fault with the forum software. But it downloads fine.

So the peak current is 234mV across 10R gives you 23.7mA. Assuming this is 20mA for the LED then the remaining 3.7mA must be the current for the chips, giving 0.62mA for each chip.

Note that is using the face value of the resistor ( 5% ) and there are 5 surface mount LEDs and one 5mm LED. So that seems to check out.

When more that one LED is on at any one time the waveform is very different. This is due to the PWM generator in each chip being free running and thus getting an odd changing waveform due to the times each LED takes a dive in current it is drawing.

Decided to continue in an other post to see if the lack of an image is fixed. EDIT - seems to be OK on this post. EDIT EDIT Spoke too soon the last one is not showing although it was before I edited the post to move the diagrams in line with the text.

So 2 LEDs on at maximum brightness, triggered to show the times the chip's PWM are in sync.

Same setup but triggered for a normal out of sync state.

Five LEDs all on full brightness.

Finally 5 LEDs all on at half brightness ( 128 ). All wave forms are just random parts of the same complex wave.

DS1Z_QuickPrint7.png

DS1Z_QuickPrint8.png

DS1Z_QuickPrint9.png

DS1Z_QuickPrint7.png

DS1Z_QuickPrint8.png

DS1Z_QuickPrint9.png

Thanks Mike for all the tests :slight_smile: So basicly, I can't rely on measurments from multimeter and when I design project with led's I just need take as granted that one led will take 60mA. I can end with more power then needed, but it will be safer than other way around.

Back to to the tests, I add extra wires to the strip (1/3, 2/3), cap and resistor, nothing changed.
I cut strip to 50 leds and connected to 5V/2A power supply, same results. With 5V 10A power supply also same results.

So bottom line, I think I can't test it more, and for now I will stick to the 60mA/led rule. Next strip I will buy from other suplier and do another tests. But probably I will get same values.

Thanks for all help!

So basicly, I can't rely on measurments from multimeter

Correct.

and when I design project with led's I just need take as granted that one led will take 60mA.

Not quite. If you set the brightness at 50% in the LED library then you can quite happily only consider 30mA per LED. Or any other brightness / current fraction. But you have to do this as a calculation rather than trying to measure the current.

Yes, that would be an option, but I don't like to limit possiblities, so I would rather go with proper power supply for max power :wink:

I have one extra question regarding wireing. I need to make some lightning tests, but maybe I will need to use 60 leds/m (or more) strips. I don't wanna use such powerfull power supply, so I consider using two power suplies for each strip.
Please advice if this will be correct:

  • first strip connected to first power supply, from this strip I will power ESP
  • second strip connected to second power supply
  • first and second strip connected only with data line

Grounds also should be connected?

Yes, of course grounds must be connected.

Using 2 psu vs 1 is not recommended if you have the choice. Variations in the output voltage can cause uneven brightness or back to the problems of level shifting.

This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.