Hey all. I have a large sculptural piece that I’d like to switch away from FastLED. It’s only 32 pixels (with some mapped gaps). The trick here is I’m using 3 x 10 channel RGB driver boards. They were originally DMX but they got damaged but I found out the boards are actually TM1812 chips with extras on top so drove them with that protocol instead.
So, I’m currently trawling the data sheets - it seems TM1812 is remarkably similar to WS1812 (probably why FastLED supports it).
Any thoughts on feasibility of using PixelBlaze to drive them?
On a quick look, it seems that each chip actually controls 4 RGB LEDs, does that match your research?
I was surprised to find that even the venerable Advatek protocol database doesn’t list these, though they do cover a TM1809 which looks related.
I think given the uncommon nature spotting these out in the wild, it’s probably unlikely we’d be building in protocol support any time soon.
Sounds like you’re not afraid of a datasheet, so if you were really deeply interested in using Pixelblaze with these, you might start from the Output Expander or other open-source protocol driver code, and work up a custom expander that can speak TM1812.
Yes the driver boards are 10xRGB for a total of 30 channels output. The boards use TM1812 chips for the step between the MPU and output MOSFETS. As the board is only 10x3 it uses 4xRGB chips hence the 2 channel gap for each board.
The datasheet for the TM chip seems to show a remarkable similarity to WS2812.
Might have a dig and see just what needs to change on that front…