I'm having a bit of a problem,

I’m having a bit of a problem, and was hoping someone might be able to help.

I have a circuit using 224 ws2812b leds, on strips of 30/meter (so 7.5 meters). This is split into 8 uneven strips (23, 23, 28, 28, 31, 32, 29, 30), run off one continuous data line for all of them, with a 300ohm resistor before the first pixel. Each strip has it’s own power connection, run off one 6A power supply. My microcontroller (lilypad usb) is powered seperately with a 2.5A lipo battery.

When I run the neopixel library teststrip program (from an uno, 5v power supply), everything works perfectly. When I switch to my program running fastLED, the final meter of leds goes crazy, intermittently. Those leds will flash red, green, blue, very quickly for a couple seconds, at full brightness, then go back to what they’re programmed to do for a few seconds, over and over (I’ve tried it at a 100 and 50 brightness value, same result both times). If I pinch any exposed part of the data line, I can force the craziness to happen.

If I bypass the resistor, the flashing on the final meter stops, but then the program runs at or near full brightness. By pinching the any exposed part of the data line then, It goes back to the programmed brightness, and the craziness happens again.

It seems I have plenty of power to run the circuit, since the teststrip program works fine, which also indicates to me everything is wired up good.

My guess is a weak data signal. Perhaps I need to run the lilypad via usb to gain a 5v signal? I’m only home on weekends and didn’t think of this until I left for work.

However, I’m confused why bypassing the resistor would cause it to run full brightness, when programmed to do much less, so I thought it best to see if anyone has any insights.

I can draw up a circuit diagram if needed, or provide code if you like.

Thank you

Are the grounds of everything, included ground of the battery powered lilypad and ground from the 6A power supply, connected together? Does it make any difference if you also power the lilypad off the 6A power supply instead of battery?

Have you measured voltage checking for 5 volts at the beginning and end of each strip?

Sure, post your code to http://gist.github.com and share a link here. And it can’t hurt to see a circuit diagram.

Yes, the grounds for everything is connected to the lilypad ground as well. I’ll post a diagram as soon as I can.

I haven’t been able to test it powering the lilypad from usb, cause I didn’t think about that until after leaving for work. Have to wait until saturday to try it. Thinking about it, seems like not having a 5v data signal would be a likely cause.

I’ll also post the code to github. It is the code you edited for me @marmil ​, so you should know it, but I’ll get it up for others.

Ok, here’s a poorly made circuit diagram done on my phone with a cheapy program, should make sense though. Note, data connections are very short, as they wrap around a skirt.

There is a 300ohm resistor on the data line, and a 1000uF capacitor across the + and - for each of the 2 6A power connections

missing/deleted image from Google+

Thank you for the drawing.
I’m confused, besides the power for the lilypad, is there one, or two 6A power supplies? If more then one can you update the diagram?

I think the Lilypad USB pins output 3.3v, correct? Maybe you need a level shifter (74HCT245) for the data line.

1 power supply, two seperate usb connections to power it (can only get up to 3A from each usb port)

That is my best guess, is that I need a 5v data signal. And you are correct, lilypad will only give 3.3v (I thought it could do 5v, but was mistaken) so yes, a logic level converter, or different board is likely needed.

I’m going to use my uno tester to see if it works, and go from there.

There might be two versions of the Lilypad, one 5v and one 3.3v.

Yeah, I have the lilypad usb, which is strictly 3.3v, but the others can operate higher (probably the cause of my mistake)

So it’ll most likely be a logic level shifter, or other board for me, lol.