WS2812B Wrong colors

[Grumpy_Mike]

Thanks for that. Yes those two boards are fakes so their might be a quality issue here.

Can you do some tests for me as I have not got your system?
With only the long strip connected reduce the green value from 150 down to 8. Now disconnect the wire from your boost converter to the 5V line of the Micro and power it from the USB socket of your computer, does it show all green now?

[3615gregoire]

Thank for your continued help Mike.

As suggested, I have:
- Disconnected the eyes strips
- Added a capacitor on the power to the LEDs strip
- Added a resistor on the Data cable
- Disconnected the power cables from the converter
- Changed the program for color: 0,8,0
- Power the system from USB


The color sequence is the same as before, at least for the 15 first LEDs (I haven't checked them further one by one).

We have done the same on my girlfriend's helmet, she as also a string of different colors. We use 2 different makes of the Mini Pro.

I'll be away from home for a week, but if you have more suggestions, I'll be happy to try more.

[Grumpy_Mike]

We use 2 different makes of the Mini Pro.

Are any of these real or are they both ripoffs?

Do you have any other sorts of Arduino you can try?

[3615gregoire]

They are both not original. I can't seem to find an original for sale online since it has been been discontinued.


I'm going to try to simplify further: reduce the length of the cable, maybe also the length of the strip.

[Grumpy_Mike]

Well I set up a Pro Micro and ran your code in reply #14. This is what I got.


So I am not seeing the same issue as you.

If I have got this right then with two strips connected, a short one and a long one, both on the same pin. Are you getting all 11 of the short strip green but the first 11 ( at least ) of the long strip different colours? That seems to suggest you have a problem with the long LED strip rather than the code or timing.

I notice from the schematic of the Pro Mini that it could be fitted with a 8, 16 or 20 MHz crystal. What is fitted on yours, it should be marked on the top.

[3615gregoire]

Sorry for the long silence, I was away from home.

I found what the issue is: it's one of the 2 kinds of LED that I purchased that creates it...
individual LEDs

I have replaced them by the other model (LED strip), and voila...

I though it was not the hardware because:
- Some from the strip also had wrong colors. I suppose the signal was polluted though the individual LEDs?
- When I swapped the 2 first individual LEDs, the wrong colors remained the same, according to their position, not to the LEDs itself.

Thank you very much for your help Mike!

[Grumpy_Mike]

Thank you very much for your help Mike!

Your very welcome, and thanks for the feedback to conclude this thread. Often people just leave things hanging, or just say it is solved without saying why.

[larsdav]

I know this is an old post - But I wanted to seal it up so people knew what the problem was.  I am 99% sure you were using RGBW (note, the forth LED, WHITE).  That will then create a chain of issues since you need to send FOUR individual codes per LED.  It chains over to the next one in line, and goes forever. 

[Pete-Repeat]

Just adding to this rather than creating a new thread. But one of my strip is showing randomly wrong colour leds, however only when below 50 percent brightness. If I fade it down then some led stay in brightly with the wrong colours. If I put full brightness and then reduce then the colours are all as expected until the brightness get low again.

This is only on one strip, my other strip with the same code is perfectly fine.

I have a 1000uF cap on the +/- supply of the strip. I have a 220 ohm resistor on the data pin of the strip. All cable lengths are reasonably short.

I believe the supply is just over 5V at the start (5.15V off memory) and very close on 5V at the end.

Should I look at my power supply, changing my resistor or looking at using a new gitial pin.

Its a buy of a pain to change things in situ which is why I haven't tried these things yet.

[Grumpy_Mike]

This is only on one strip, my other strip with the same code is perfectly fine.

Look to see if this strip has a ceramic capacitor on every LED. I once got a strip that only had them fitted on every eighth LED.

[Pete-Repeat]

I cut a 5m strip in half at the cut marks and made thr two strip's. So each strip is the same, other than the one working 100 percent running off a lab power supply with no capacitor or resistors yet, and the one not working right when dimmed down running off a plug pack however has the capacitor and resistor in it... I had hoped addimg the capacitor and resistor would fix it but it made no difference to random wrong pixel colour (or individual pixels night when the rest are off).

I will check about the capacitors on the strip however.

As the led work fine at 50 percent and upwards in brightness I was thinking that the strip isn't damaged and it's not to do with voltage drop. But this is an uneducated stab at the problem.

[Grumpy_Mike]

and the one not working right when dimmed down running off a plug pack

I am assuming that a plug pack is some sort of power supply. I am also assuming that the ground of this plug pack is also connected to the ground of the Arduino and the +ve is only connected to this second strip.
Is this correct?

running off a lab power supply with no capacitor or resistors yet,

Well put them in as soon as.

[Paul__B]

I am assuming that a plug pack is some sort of power supply.