Not specifically a FastLED problem but I thought I'd post here since you guys

Archive FastLED Archive
,
1

Not specifically a FastLED problem but I thought I’d post here since you guys have the best experience with LEDs.

I’m running a really long strip of APA102 LEDs. A total of 1758 leds in a single line. These were soldered together from 13x 5meter strips into a 52x34 LED matrix.

They are working absolutely perfectly up until about the 15th row, at which point it starts flickering. Looks like data corruption to me, since if I disconnect the data line the LEDs stay static so I don’t think its a power issue.

These APA102 strips seem to have a cap and a resistor at the start of every 5meter strip. I kept these in when joining the strips as I figured they wouldn’t do any harm. Could this potentially be the problem?

Or is a data line that long just not going to work? Is there perhaps a way I can reduce the noise on the data / clock lines?

I’m uploading a video now to demonstrate the problem, but any thoughts would be much appreciated.

2

A couple of things - where are you injecting power? You need to be doing this every hundred leds or so.

You can also try scaling the speed that SPI data is being written out back a bit:

LEDS.addLeds<APA102,RGB,DATA_RATE_MHZ(12)>(leds, NUM_LEDS);

Right now it is set to DATA_RATE_MHZ(24) I believe.

3

Hey Daniel,

I’m injecting power and ground at every row, so I think thats ok.

Turns out you were right, turned the SPI down to 3mhz and its working! Thanks very much Daniel.

4

@Daniel_Garcia Is the data rate supposed to decrease, as the number of LEDs increase? I was able to get a DATA_RATE_MHZ(14) without flicker on just over 100 LEDs.

5

Not necessarily. I think it is a function of wiring quality/power, because I have a matrix of 768 apa102s that I can run at 24Mhz.

6

I am running 1758 LEDs. But I am driving them from the PRU on a beaglebone black so not actually using FastLED. I may have also misconfigured the registers so I will have to check again when I have a brain that isn’t addled by solder fumes. :wink:

7

Does the badness occur at a point where one strip length is soldered to another by the factory? Or if you press on the strip around the bad point, do things change?

I’ve had problems with bad factory solder joints both between strip lengths and on an LED near a length edge. And these bad joints only became noticeable when the strip was unrolled/installed.