I am using the command-line image viewer and video viewer to display images and videos on a 2 x 2 matrix of P3 64x64 matrices that I ordered from electrodragon along with their HUB75 hzeller Pi shield. I connected P4 to GND and P8 zu #E for using the 64 x 64 panels. My images and animations all contain colourful objects on a black background. When I display them, I always get some random horizontal linear pixel glitches/artefacts next to the outline of the shapes I am drawing as visible in the photo below. These are also visible when running the demo D0. I have tried different panels from amazon and different panel sizes (1x2, 1x1) but no change. I also tried all options discussed under the “ghosting” issue in the readme to no avail. Does anybody have an idea what might be causing this?
battery: do you mean the old electrodragon active-3 board? If so, it’s for an optional RTC on the board. You can likely ignore that.
ghosting: I don’t see this but the best things I can think about are
change the ribbon cables, get better ones, get shorter length
make sure your power ground between rPi and panels is the same
sadly that’s not easy to debug, but I’m not seeing this on any of my panels or boards.
Thanks for the power ground tip, panels and Pi are currently not on same ground and I had grounding issues with audio equipment in that room before so thats def a possibility, will try that and report back.
ok, good news on grounding, that is most likely your issue.
RGB Matrix Panel Drive Board For Raspberry Pi V2 - ElectroDragon is the new version of the board, it just simplifies the board, removes the RTC almost no one used, and adds angle connectors to make the profile a bit more flat.
You don’t need to change boards, yours is functionally equivalent.
Thanks for the link to the new Pi board, I hadn’t seen that one!
Regarding the glitches/artefacts, I had no luck with connecting the RP to the LED panels’ power source. I tried with another batch of LED panels (leaving everything else the same, including the wiring) and voila, the glitches are gone. So it seems to be an issue specific to the panels.
Oddly, I ordered the second working batch from electrodragon just like I ordered the first batch with the glitches a few weeks prior so I was expecting the same panels. They are clearly different though as their backs look different. Is there a way to find out whether they have a different chipset or how they differ so that I can avoid the “bad” ones when ordering another batch?
Thanks again for the help and the awesome package and happy holidays!
unfortunately you just found out the problem with RGB Panels: unless you have a personal contact/salesperson in China who can guarantee certain panels, you will get the panels of the day, and in some extreme cases, the panels they send are totally incompatible (like ABC instead of ABCDE)
Haha ok thanks, this turns out to be a fun hobby. The panels with the glitches all have the FM6124DJ driver. Can you confirm that generally, panels with this driver should work? For what its worth, the panels that work fine have the DP5125D driver.
Would it make sense to assemble a list of tested/working drivers (and potentially where the panel came from) here or on github?
I don’t think anyone has made a list of panels and chips that works, and it’s not a bad idea.
What I recommend is you make a new thread saying “please report working and non working chips”, and everyone can contribute with list of what works and what doesn’t. After enough time has passed and enough answers have been given, you can submit a github request to add that info to the main page (best to do it all at once at start, and then later, we can make updates as time goes by and new chip info gets added)
Although the table contains panel compatibility data for another library, panels marked in green in the table will probably work on most libraries. (excluding a few S-PWM type drivers whose support is exclusive to the DMD_STM32 library).