I have found that testing the variables to be invaluable with WS2812B leds - the can be tricky little bastards
also .. how is the USB being split? are their multiple FC controllers?
If you are running through a hub - detach and go straight to the machine
the key is to determine for certain it is not the first pixel in the chain of LEDs
Can you swap that channel for another strip and it work correctly?
Just one FC, plugged directly into a rpi2. I can take known good strips and plug them into the first couple of channels (with shorter runs) and they work fine, but when I put them on the end of the long run, don't work. So, effectively did the 'bring the FC closer' except I brought the strip closer. ;)
did you solder header pins to your FC?
and if so - do they all share a common ground?
at one point I made the mistake of splitting the grounds between two power supplies and it did not like that
You maybe at a physical limitation for the FC and transmission over a long distance
it is a known limitation unfortunately
I did solder header pins. Right now I've only got one of the grounds connected, since they're all tied together on the FC anyhow (or so says the schematic). I wonder if the half-inch difference in return path could actually cause that sort of issue. Would surprise me.
I had a 3 meter run that I had to do recently on another project and ended using a 10' extension USB cable and it fixed my problem
you are correct they are all common on the board
I should have noted that in my situation where the grounds were split, it was not at the FC - it was at the strip where I jumped it back to the bus for all the negative terminals - I had the strip on one PS ground and the FC on a separate one - I suspect this in some way messed with the timing or something - once I had common grounds across the whole display, all of the strange flickering in that situation went away
Wish I had more to offer :(
That's odd ... same power supply, but different ground screws? (Like one of the typical-ish 5v/60A supplies w/ three grounds and three +5v?)
I am actually using 2801 LEDs for my next display due to the limitation you are hitting right now - not giving up on FC though as it has some really awesome applications - though I do wish there were a headless all in one option, but that was not part of the original design motivation I don't think - I am currently (tonight) messing with trying to get the ARTNET functionality to work with http://www.solderlab.de/index.php/software/glediator - just now reading through all of the code
Define: "headless all in one"?
meaning a controller with something like the RPzero or small form factor arduino
one thing I like about the arduino for the signs that I have lit lately is that running my configs without the FC and straight from an Arduino is much more reliable from a power on power off perspective - RP as you probably know have a huge issue with power loss - which means in addition to powering something installed for a long time PLUS providing a battery backup that shuts down the display in the event of a power failure becomes painful
Using the FASTLED libraries on an Ardiuno has become my goto for all small projects - the FC does excel for large scale projects - particularly the ingenious implementation of OPC and all of @scanlime goodness.
Much different foot print for sure - but has a lot of built in components that make doing large installs much easier it would seem - would be awesome to see where FC could be integrated into a project like this
buy this board - snap in a FC and away you go with all the ports and connections ready to roll