colorlight 5a-75b not working with raspberry pi4B fpp

[aquila2heman]

Hi All, 
I am doing a project with 48 P5 32X32 RGB panels in a single row display with 3 Colorlight 5a-75b cards, connecting 4 panels Desi chains per 1slot and 6 slots per Colorlight card. I am trying to connect the system to FPP Raspberry Pi4 B. It works well with LED Vision but won't work with FPP. I have done all the setup according to YouTube videos. 

The Green LED on colorlight card is starting to blink when i run FPP test program, but the display doesn't do anything.

I tried connecting to just 4 panels and setting the settings accordingly in LED Vision and FPP, but still it didn't work for FPP.  

I hope any of you experts can guide me to get the results I wanted. 

Thanks 

[JonB256]

I use a PiB to drive a single Colorlight connected to 48 P5 panels (six rows of eight). It works flawlessly.
My panels are 64x32 so not the same as yours. You should post screen shots of your FPP Output screen.

[aquila2heman]

Thanks for the reply, I tried with 4 panels with 32x32 x4 on 2 slots on Color light card still it doesn't work with FPP.
 You cannot view this attachment.

[Kensington Graves]

Colorlight receiver cards have a FPGA chip that LEDVision flashes with information about the receiver's place in the overall matrix.  This means that you can't just break apart the matrix into smaller pieces and have it work.  You would need to reconfigure the matrix as a 2x2 and then reflash the board with LEDVision to get the smaller matrix to work.

Go back to LEDVision and reconfigure for the smaller matrix.  Then go to FPP, choose the panel size and scan rate, and use the "auto layout" button to set the output/panel number instead of manually inputting that information.  The "auto layout" button resulted from CaptainMurdoch troubleshooting my matrix last year and simplifies the process of mapping multiple Colorlights in large matrices.  I found that my output/panel numbers were different from what I had manually selected (which selections resulted in wonky output).

[JonB256]

You have the output set to eth0 from the Pi. Is it possible that you have an eth1 connection for a USB Gigabit adapter that you should be using? (I do)

[aquila2heman]

Go back to LEDVision and reconfigure for the smaller matrix.  Then go to FPP, choose the panel size and scan rate, and use the "auto layout" button to set the output/panel number instead of manually inputting that information.

Thank you for the reply,

I did test as you mentioned already and still it works with LED Vision but not with FPP. I'm doubting now if my new Raspberry Pi 4 is having the fault with communicating with Colorlight cards.

Also, I found when I am testing with FPP if I remove the power from the Colorlight card and replugged it, it will flash the panels with correct test display but just for less than a second.

Thanks

[aquila2heman]

You have the output set to eth0 from the Pi. Is it possible that you have an eth1 connection for a USB Gigabit adapter that you should be using? (I do)

Thank you for the reply,

I already tried with that too. Still didn't get the results. Anyway, I'm ordering one more new gigabyte Ethernet adapter to test again.

I'm doubting now if my new Raspberry Pi 4 is having the fault with communicating with Colorlight cards.

Also, I found when I am testing with FPP if I remove the power from the Colorlight card and replugged it, it will flash the panels with correct test display but just for less than a second then again blank.

Thanks
 

[JonB256]

I don't see anyplace where you tried a different cable. Your CAT5e cable from PC/LEDVision must be good. Is that the same cable? When you mention that it flashed the correct pattern briefly, that occurred to me.

[Kensington Graves]

I think Jon is probably on the right track.  Generally, if LEDVision displays the test pattern properly but the panels don't work on FPP, something in the FPP configuration is incorrect or there's a problem with the network connection.  FPP "speaks" the Colorlight and Linsn protocols.

Have you tried changing the scan rate in FPP?  Though rare, some of the 32x32 P5s can be 1/8 scan, particularly if you have outdoor panels.  I couldn't get my P10s to work on FPP until I learned they were 1/4 scan.

Also, check your pixel overlay models in FPP to make sure the correct model is being used for the test.  You may have multiple models that have piled up during testing an you may want to clean it up to make sure you're sending data to the right model.

And check your ribbon cable connections.  I can't tell you how many times a problem was caused by a partially seated cable—like the panel would work then would not.

Maybe test the voltage from the PSU driving the panels, too.  Low or high voltage can produce odd results with panels, in my experience.

Post a screenshot of the LEDVision settings for the panels and I'll compare them to what I use.  Also, post a picture of the back of one of the panels.  That might have some helpful information on it.

[aquila2heman]

Still It same, 
Please see the images. 

Thanks

[aquila2heman]

And ThisYou cannot view this attachment.

[aquila2heman]

And this

[darylc]

And ThisYou cannot view this attachment.

This is not correct, the text is overlapping.  Until you get this part sorted, forget about FPP.

[JonB256]

What does your LEDVision Receiver Mapping screen look like?  I don't see any glaring issues on the one you showed earlier.

[MikeKrebs]

I don't know much about the panels but your led screen settings seem to say the panels are 32x16. In fpp, it says they are 32x32. Are the modules linked together to mimic 32x32 by using two 32x16 panel layouts?