Problem with power supplies

Hi,

I have an array of 140 WS2812B leds and I have 3 different powers supplies:

1 - a 5V breadboard power supply powered by a 9V 2A wall wart. 2 - a 5V 2.5A wall wart 3 - a 5V 10A wall wart

When used on the LEDs, #1 works fine but I am having issues with #2 and #3. When these are used the LEDs do not act as expected, not all turn on and the colours are random.

I have tried different amounts of LEDs; 10, 20, 40 and I get the same results as above.

What may be the problem with #2 and #3 power supplies? What kind of PSU should I be looking for to power RGB LEDs?

godivaPrima:
Have you measured the voltages from those PSUs?

Yes, 2 and 3 all are about 5V; 5.1V and 5.3V.

Are they clean though? Spikes and noise could be an issue if they are cheap (don't buy cheap mains supplies basically!)

Add 500uF or 1000uF across the supply output and see if that improves matters.

MarkT:
Are they clean though? Spikes and noise could be an issue if they are cheap (don’t buy cheap
mains supplies basically!)

Add 500uF or 1000uF across the supply output and see if that improves matters.

I already have two 470uF capacitors in parallel across vcc & gnd. 470uF are the largest I have.

Is there a way to tell if the PSUs are noisy without an oscilloscope?

Not sure if it helps but the 10A PSU is a YU0510 the same as this one.
I have two of these and both give similar results.

The smaller one is model PSC11R-05 branded Logitech and labelled as a switching Power supply (Just noticed it is 2A not 2.5A as stated above).

you have to measure the voltages from those PSUs when it is connected to the LEDs (it supply to LEDs) and ground is connected to arduino.

It is possible that the power supply switching frequency (or its harmonics) is exciting a resonance in your power distribution. Try to add a bypass (.1uF poly or ceramic) at a few random places along the power wires. A ferrite can also eat the resonance energy.

Thanks for the suggestions. I will give it another go this weekend.

9V X 2A = 18 Watts; 5V X 2.5A = 12.5 Watts; 5V X 10A = 50 Watts

and

load total 5V X 60 mA X 140 nos = 42 watts; load PSU as per wattage.

MartynC: Is there a way to tell if the PSUs are noisy without an oscilloscope?

Not really - you could capacitively couple it to a loudspeaker, but that only lets you hear the noise in the audio spectrum, and there'd by a loud crack/thump as the cap charged up which might not do the speaker much good.

Low frequency issues can be seen on a multimeter on AC volts range, might be worth a check.

Can you post all three oscilloscope traces at 1V/div 10ms/div DC coupled.
I’ll bet that the ones which don’t work are less flat than the one which does work.
I could be wrong, in which case it will be the spikes mentioned by ron_sutherland, which might need a faster timebase to see.

Tried adding bigger capacitors and also tried a very basic filter but still had the same issues.

I don't know what is wrong with the 2 PSUs and I don't have the tools to investigate so I purchased a more expensive power supply which is working OK.

.