Using the P2 to drive a really long string of pixels

[JonB256]

The P2 specs sound very useful. Being able to drive 4 universes of pixels from one output (with the necessary power injection along the way) can really reduce the number of controllers for a matrix of pixels.


But - since don't have a P2 and am just now buying pixels for a matrix, I'm wondering if anyone knows from experience if you would be able to see the physical time delay from the First pixel to the Last pixel if you really added 4 universes (4x170 pixels)? How long does it actually take to pass the data thru 680 pixels?


One of the Horizontal Matrix plans I've sketched would actually need both P2 outputs to drive the Top Half and the Bottom Half of the matrix. At the middle of the matrix, where the second output connects, would the average viewer see it Start slightly ahead of the row above it (from the first output)? I'm sure that driving the pixels at the fastest possible refresh clock would help, but am curious.

 

[j1sys]

it depends on pixel type and drive speed.

for example: 680 pixel of WS2801 at 1mbps = 16ms (680 x 24 / 1000000)

whether that would be perceptible is subject to discussion. other pixels types and speed would be different. the popular WS2811 is actually slower in real life due to the way we have to output 5 bits per bit to form the assymetric timing.

TANSTAAFL - There ain't no such thing as a free lunch

we designed these units with flexible outputs. a lot of people needed to run strings longer than a single universe. we allowed for a maximum of 4 universes since the WS2801 speed would still be below the 25ms timing that is closely tied to the DMX world. However I personally do not recommend full 4 universe strings. The timing and single point of failure starts to become a risk as you go over 300-400 pixels. we are really targeting a more realistic limit somewhere in that range.

-Ed

 

[JonB256]

WS2811 is actually slower in real life due to the way we have to output 5 bits per bit to form the assymetric timing.


-Ed

Is this due to the combined Data/Clock line? (3 wire vs. 4 wire)


I am leery of putting so many pixels onto one controller output. Was thinking of limiting length to 120 or 160 since my horizontal bars are going be 40 pixels long. The possibility of lag occurred to me as I considered the very attractive option of running the entire matrix from a single ECG-P2. (40 x 24 which would be 40 x 12 per P2 output)