(WORK AROUND) WS2815 Ghost Pixel - Help Needed

[Nojoy]

Have setup two 750 pixel Matrices using WS2815 Strip Lights.
50 per row * 15 rows.
On one of the Matrix, I have this really odd issue where pixel 346 is not being accounted for.

When I run a sequence, Pixel 345 moves straight to 347. I've checked:

1. Traces and solder on pixel and cuts - no issues identified.
2. Pixel 346 (channels 3130-3132) are shown in both Xlights and in the Controller.
3. There are no channel conflicts, and I have more Universes allocated than filled by the props.
4. When I Test the pixels in Xlights Pixel 345 lights correctly (and each channel tests correctly).
5. When I test Pixel 347, the next Pixel after 345 lights up correctly.
6. When I test Pixel 346, nothing happens.
7. When I test each channel of PIxel 346 while Pixel 345 is also lit up, Pixel 345 starts flashing.

Any ideas?

 

[Nojoy]

When you say 346 is dead though, there is physically no 346. 345 (counting from start) lights up appropriately, and the next pixel after 345 also lights up appropriately when 347 is selected.

 

[Grozzy]

Is your universe size mismatched somewhere 512 vs 510. What controller are you using and have you tried testing directly from the controller.

 

[TerryK]

At first, I would suggest as Kate had that pixel 346 is defective. But your post #3 indicates there is no pixel 346. But if 347 works then the pixel after 345 works. So what pixel follows 345?

I first suggest if you have 2 matrices and if they are identical as you appear to indicate, that you swap them and see if the problem follows the matrix or stays with the controller or controller port. Obviously if the problem moves it is the matrix, if not either or both the controller or software configuration.

Your 1st post item 2 mentions pixel 346 is channels 3130-3132. Is pixel 345 channels 3127-3129 and pixel 347 channels 3133-3135? I too suggest using xLights for testing. I would turn on only channel 3127 and using the drive level 'slider' shove that channel's binary drive level to 255 and back to 0. Look for a smooth illumination of the correct color. Any jumps in brightness or flashes of mixed colors indicates that pixel is bad. Turn that channel off, move to the next and repeat.

 

[Nojoy]

Hi all, life got in way of my troubleshooting but appreciate the responses.
*Checked universes, no conflicts
*plugged the second matrix into the same output, renders and displays perfectly so have ruled out controller, universe or channel issues
* have cut and resoldered the connection from 345 to the next pixel, issue remains
Issue being that 345 lights correctly and has no issue when tested from within xlights or directly from FPP, but when 346 is tested nothing happens. When 347 is tested, the pixel after 345 reacts to the 347 test inputs.
* have developed a work around by re-creating the matrix using single polylines, and turning 346 into a null pixel. Grouped all into a model and now displaying effects without the issue. Hasn't resolved the ghost 346, but I can work around it. Just hope any sequences I share with UXSG are able to be mapped by other users without too much issue (not sure if they just see my 'Matrix Left' group or all the submodels).