I'd quite sure that this will work (giving my reasonably competent knowledge of protocols), but wanting to know if others have done this.
I have a Pentagram in coroplast that needs a little more light in each segment, so was thinking of parallelising up the 3-LED rectangle pixel blocks, so that it doesn't increase the channel count. (I cannot think of a reason why I would want to not have the same segment with the same color.)
A couple of points on splitting pixel data, Normally a Y connection in a data line is a huge no-no. However, there are exceptions.
A Y with any long length of wire (more than 8 inches) will cause problems. Pixels wired in parallel with very short leads is possible and the pixels at the connection point will all have the same address. An example where this can be used is LED fireworks. A strip up a tree can simulate the rocket, and the end can branch out to say 4 pixels, go a few pixels, branch again, and continue to simulate a shell burst. Each ring of pixes will have the same address. If you can keep the distance from the split to the pixels to under 2 inches, it may work for you.
A Y in a data line makes an electrical device called a stub. Like an antenna, it is an electrical device to the frequencies of the data. This stub will reflect data back into the data line and corrupt the data. You can't T or Y a data line with wire going off to other locations. Don't even try it. Driving 4 pixels together at the end of the data wire is legal. Keep the wire very very short as if the pixels are on the same printed circuit board next to each other.
As far as the additional electrical load on the data line due to the parallel load, I found the WS2811 pixels I am using are high impedance and draw almost no current so splitting 4 ways is verified to work. Remember,wire between the split and pixel must be very short for it to work.
In regular electronics such as TTL, this capacity to split is referred to FAN OUT. 1 chip driving other chips has enough current to drive the imput of several other chips. I have never seen an officail fan out spec for pixels. In normal design the fan out is 1, but I do know it is more than one, but I don't know the max.
I made a Y being ignorant to any counter information and it worked with two Valentine's Day Hearts made up of 25 GECE's each. I had one output split to two sets. It worked fine. Can't say if you have the same results as it's a bit different.
The Y split, to the first pixel will be within 2inches, and then only size pixels after that..
Thanks for the advice, I'll report my success here.
/BM
scootchu wrote:I made a Y being ignorant to any counter information and it worked with two Valentine's Day Hearts made up of 25 GECE's each. I had one output split to two sets. It worked fine. Can't say if you have the same results as it's a bit different.
I did the same with 2 strings of 25 GECE's. 15' to a null (but live) pixel, another 15', then split the data.
