This is moving faster than I had expected. I've posted questions here before and not gotten a lot of traction, but this one, people are very gracious to help. It is only January 5, I have a day job, and 90% of my 2021 show is still online. One longshot possibility is, I might be able to pull off an IC. People are willing to buy this (me included), especially if you break out a frame or three and route to a triac. My friend actually patented an on-IC 120vac LED (needless to say, you have to widen and spread the wires just a little bit when you run 120vac into an IC). A combined splitter/pix2ac is basically a Swiss army knife board, and could be transformative if the price is right. That's where I stand. But 1/5/2022 was a bit sooner than I had planned to lower nose to the grindstone.
Incorporating pix2ac support doesn't really fit very well. As far as I can tell, most of the pix2ac devices are based on using the LED-driver pins of the ws28xx parts to control a relay. Since there isn't any ws28xx part in the design (so far), that doesn't fit. And I'm not very keen on decoding the pixel data stream inside the parts under consideration, and bringing out a decoded signal only makes sense to me in a charleskerr or BarnabyBear device (i.e. a multiple-output splitter placed in a protected location).
I don't expect much progress in the next few weeks, at least not on my part. My son is flying in and wants to do some ice fishing, although it will just be local. For now I'm just going to sit back and see how ideas gell.
Yes, I was thinking in terms of the BBear. Or even more specifically, a breadboard version of it. If we begin with a high-level system description:
- external chip counts number of rising edges and decrements the counter whenever there is one
- when counter reaches 0, raise an interrupt pin.
- forward all subsequent ws2811 incoming packets until there is no longer a rising edge within the timing window
- repeat
By simply injecting other counters and adding another interrupt and/or if/then clauses in the microcode, you can assign frame numbers to SSR, triacs, or dumb RGB all day. Or at least, until you have to code an ISR that takes longer than 1us to run. And branch-if instructions don't take that long to execute. The SSR would not need a dedicated ws2811 IC--it needs only tie up another GPIO pin. Although you would have to include the SSR itself in your hardware design. The triac and dumb RGB don't necessarily need a ws2811 IC, either--what they really need are off-board PWM drivers. That driver could be a ws2811 IC.
A DIY'er could go to town with that on a breadboard. But in order to get a fair price from someone who just wants a board, you have to freeze in one architecture at a time (as you are aware--you just designed jumpers). Economically, being able to buy a cheaper dumb RGB strip offsets the price of the board. You're not just asking people for more money: instead you are offering them a tradeoff. If they don't need the individual pixel control, instead they can buy the board+dumb RGB for the same price as pixels, and now they get this other capability with port splitting, AC, a built-in data repeater, etc..
Hi, for me it was easier to have a standalone WS2811 triac module that could be fitted anywhere in a WS2811 data stream.
https://diychristmas.org/vb5/node/9664
Thanks again. My logic is that the more you combine your boards, the less weatherproofing you have to deal with. Voltage drop is a non-issue when routing AC. The board would locate along a long pixel data line, but then you can tap it to run AC. That location tends to be in the thick of things (and is guaranteed to be in the thick of things, pixel-wise), so the AC line would probably not normally be that long. By consolidating multi-function, you reduce the number of boards, reduce weatherproofing, reduce aesthetic issues. Possibly reduce the number of differential receivers.
