Oh, right—you mean is there a potential difference between ground at the power supplies and ground on the strips? I measured that at one point and it’s some small number of mV, it didn’t seem significant to me.
As @Daniel_Garcia tested the code with 60 leds per channel have you tried it with your software limited to 60 per channel to see if the problems are still visible?
I cut it down to 80/channel and still had the same problems—it was fiddly to try 60 because reasons, so I didn’t, but I’ll give that a go tomorrow.
Yes this is what I would have expected a larger difference between GND at the PSU’s and GND at the beginning and the end of the strips.
What size wire carries the DC volts to the strips ?
what is that length ?
What is the PSU’s DC volts out ?
1.5mm^2, 4m, 12.15V.
Conversion for you imperialists:
1.5mm^2 = slightly thicker than 16AWG
4m = 13.12ft
12.15v = 12.15v
Not sure the current and voltage drop, but I just want to point out that I always found this page very helpful for figuring out current, wire sizing and voltage drops:
http://www.powerstream.com/Wire_Size.htm
(It could be better… like giving 5v as an option in the drop down list, but its still nice).
carry on,
-frenchy (Steve French)
http://www.voltvision.com
@Robert_Atkins I have done a lot of my rigs using DC-DC stepdown converters. I recently purchased a heavy duty 20amp model that had a sweet heatsink that was exposed through the plastic housing. When I had a 16x16 WS2812 matrix set to full brightness, whenever I would accidentally graze across the heatsink with my skin, I would get the same exact results from my LEDs that I see in your strips.
I am not sure what exactly is going on, but I can only guess that the DC converter wasn’t supplying the correct amount of power or possibly an unusual waveform in the current to the LEDs. IS there possibly something going on with the DC converters you are using and the orientation/setup they are in that causes this activity when a certain power draw is reached?
That’s the standard suspicion @Jon_Burroughs , “if there’s something funny going on, your LEDs aren’t getting enough power” but the couple of sticks I’ve added GBFH* caps on still have the problem. I’m well under the rated power output of the DC-DC converters in this configuration, and these aren’t eBay cheapies with “aspirational” specs, these are pro units with proper datasheets (http://power.murata.com/data/power/okr-t10-w12.pdf).
(*GBFH == “very large” aka 1000µF)
@Robert_Atkins I just came back to this thread, and see you added a point about it working until it warms up, and or you increase the load through adding more LEDs. The combination of those two things would have me looking at the power.
Get a can of freeze spray, or invert a can of “air” ( the type used for blowing shit out of equipment) and hit the DC-DC converters one by one, and the main PSU, and see if a cold blast “fixes” it temporarily. If so, you got some thermal problems affecting your power.
Also given the “dodgy pins are always dodgy”, triple check that wiring.
How many LEDs is a murata (I don’t really suspect them) powering?
and try doubling up the GND wire to an affected stick.
So, for future thread archeologists, the solution was to turn overclocking off on the Teensy. I did that and it’s all 100%: https://plus.google.com/u/0/112974620627169436992/posts/iX8hwaayycT