Led Strip WS2812B Bug (the start is turn on)

Hello,
I currently have a problem with a WS2812B led strip.
I think it's a physical problem because no matter what program I always have the same bug.
I have the beginning of my led strip which stays on.
More precisely, it lights up as long as no order is given.
When I give him the order to light an LED, this part will go out.
And as soon as my led made the order ask. The beginning part lights up.
Do you know what can cause this problem.
FYI: My power supply for my 54 LEDs is 5 V - 1.2A (I tried with more intensity, it didn't change anything)
I use the PIN 10 of an arduino mega for the control of the led (with a resistance of 330 Ohm in series (without this resistance i have the same problem))
The Gnds of my card and the LED power supply are connected.

Can you Help Me

Thank you very much

What about the other LEDs? Do they function normally?

I Have 54 Leds the 16 First is alwas white the 17th leds is yellow and the other works normaly

Finnaly no the number are false the number of no-working leds are random (but aproximately 16)
this is my light strip when a want it turn off and the 2nd picture is when a want it turn on blue

Not sure about FastLed library, but the Adafruit library has timing settings that you sometimes have to change for different "flavours" of WS2812/WS2812B and newer variants. With parts from China, you sometimes don't know exactly which kind of NeoPixel you're getting.

Are you sure you're feeding the right end of the strip? What kind of wiring? Would it be possible to take images of your hardware setup and post them here?

melvinlebailly:
it lights up as long as no order is given.
When I give him the order to light an LED, this part will go out.

So you are saying some LEDs light up randomly before the Arduino sends any data? This is normal.

melvinlebailly:
And as soon as my led made the order ask.

I can't understand you.

PaulRB:
So you are saying some LEDs light up randomly before the Arduino sends any data? This is normal.

Not his present problem by the sounds of it, but a pull-down on the Arduino output pin should largely suppress that.

Hi,

Can you please post a copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png?

Can you please post your code, using code tags?

What model Arduino are you using?

Tom... :slight_smile:

Paul__B:
a pull-down on the Arduino output pin should largely suppress that.

In case of random signals on the data line which is floating as the Arduino boots? I suppose a pull-down would fix that, but don't most of these LEDs start up in an undefined state anyway, until data is received?

PaulRB:
but don’t most of these LEDs start up in an undefined state anyway, until data is received?

Not according to bullet 5 on the datasheet! :sunglasses:

Paul__B:
Not according to bullet 5 on the datasheet! :sunglasses:

Its not very clear. That’s the only place its mentioned, there’s no further explanation of exactly what “electric reset circuit” means or does elsewhere in the data sheet. It does not explicitly say that all registers are set to zero. That “electric reset circuit” might only put the chip into a state where its ready to respond to data, rather than getting stuck in some “hung” state, but not actually clear the registers to zero.

Plus we don’t know if the OP’s leds were manufactured by Worldsemi. They might be knock-off chips.

In the absence of commands, all the NeoPixels I have, always remain off at power up. But if the input is high impedance, they will often light up randomly. I know this from instances where the connections were intermittently open. So all my NeoPixel sketches initialize the drive pin to output as soon as possible.

Actually, it crossed my mind that this might be happening in the photos that were posted above.

I have a project that drives 256 Neo's from an UNO. I depend on them remaining off at boot time because they can only be operated at very low intensity that way. If more than a few of them came on at full brightness, it would kill the UNO power and reset it. The way I have it set up, it always works.

There is probably enough capacitance on the data line that it remains low for a split second at boot time, just enough time to initialize the output pin.

Easier to simply fit the pull-down, I would say. :roll_eyes: 10k should be perfect.