Working on a project for my Burning Man camp and I am running into

fastled-support
gplus
(Sean Winn) #1

Working on a project for my Burning Man camp and I am running into the same problems on this and another project. The project uses a 5v 60A power supply and 11 ws2812b strips with an arduino nano powering everything. When powering everything up, sometimes it works - sometimes it does not. Re-plugging the 3-pin connectors usually fixes the issue.

I have nothing inline here. No caps. No level shifters. Ideas and advice on how I can troubleshoot and fix this in the coming weeks is greatly appreciated.

(Andrew Tuline) #2

Just a wild guess, but it sounds like your code may be referring to leds[NUM_LEDS] at some point.

The last location you should be addressing is leds[NUM_LEDS-1].

(Ben Schattinger) #3

I’m assuming that you’re using a 5V Nano. How is your Arduino Nano powered? Do you have the power supply’s positive and negative leads connected to both the start of the LEDs and the Arduino? Also, do you have any spare big capacitors? I’d try a 1000µF capacitor as close to the strip as possible.

(Ken White) #4

@Sean_Winn - I assume that you have more than 255 RGB LEDs. If so, are any of your variables that can go above 255 defined as uint8_t? This will cause problems. You will need to define those variables as uint16_t or a higher variable range.

(Ben Schattinger) #5

@Andrew_Tuline @Ken_White The reason that I don’t think that that is the issue is because he said that it only happens some of the time, and replugging the “3-pin connector” (not sure if that is just for the LEDs, or powers the Arduino too) would point to an issue with the LEDs. Could still be that, though.

(Sean Winn) #6

@Ken_White All of the strands are wired in parallel such that there is no more than 150 LEDS in length on any single strand.

To clarify, yhe 3-pin connectors refer just to the JST connector for the LED strands themselves. After powering on the PS, the strands will occasionally light up using the programmed pattern. Other times it will be stuck with some random lights that are lit using what I can only assume are stored values. If I unplug the 3-pin JST connector between the LEDS and the PS/Data Pin - it will start to display the correct program.

(Sean Winn) #7

Forgot to add that I re-connect the JST wires and THEN it starts to work. This is not magic.

(Marc Miller) #8

I would guess this is a hardware issue and not code related (especially if you still see this behavior when running the FastLED examples that are known to work).

How long is the data line from the Nano to the first strip?
Maybe try adding a resistor to the data line and 1000uF capacitor?

What happens if you power up your Nano a few seconds AFTER powering the strips? Or reset the Nano if it’s not displaying correctly? Or try adding a two second delay at the start of your setup() section? Does any of that cause it to run correctly?

Btw, that display reminds me of an Imperial Star Destroyer in an alternate techno universe. Which I love the idea of. :slight_smile:

(Yves BAZIN) #9

@Sean_Winn can you show a diagram of your build? And stupid question if you first powered your led without any signal do they light up white or stay black ?

(Randal B) #10

Capacitor and making sure you have a slight delay on the startup of your code usually does the trick for me.