Possible power supply issue causing weird pixel issue

[LawrenceDriveLights]

Hi guys,

wasn't sure where to post this if in the wrong section please move etc.

I am currently playing around with my christmas lights (as you do in June!) and I am having a weird issue with one pixel staying on red all the time.

Quick background is using a Falcon F48 pixel controller powered by a generic Chinese 5 volt power supply (30 amps or something like that, as it was the only 5 volt power supply I had sitting around to use).

Using a differential receiver board with 12 volt Ws2811 pixels.

Issue is one of the pixels at the start if the string (on a pixel tree with 3 x strings of 168 pixels per string and power injecting via the last pixel connection) is staying on red all the time and when running a test sequence it does crazy stuff on this string.

I tried attaching a known good pixel to the output from the differential receiver and it still stayed on red (eliminating a possible pixel issue), i even swapped the strings around and the problem followed the same output from the diff receiver board.

I tried swapping around pigtails and leads, nope pixel still staying on red.

I tried removing the power injection, nope pixel still staying on red.

I tried separating the cat cables from the F48 to the differential receiver board, nope still staying red.

I tried powering the 12 volt power supply for the problem area from a different 240 v circuit, nope still staying red.

I tried swapping over differential receiver boards, nope still staying on red.

I was starting to go a bit crazy at this point on time, so for some reason I tried swapping over the 5 volt power supply for the Falcon F48. What the????? no more pixel staying on.

Here is a quick (5 min) video to try and show you guys.

I am a bit lost, any ideas why except possible weird frequency caused by the chinese 5 volt power supply?

Please note the alternative 5 volt power supply is just a temp (Its a 4 amps USB charger), yep a bit dodgy but it was the only other 5 volt power supply I had sitting around.

Do you guys have any other ideas???? It all seems a bit weird to me.


View: https://www.youtube.com/watch?v=rdSQav-Y544&t=49s

 

[Derf]

Possibly Rouge Radio Frequencies?

Bit of a wild shot but you seem to have gone through just about every troubleshooting step. It could be possible RF interference from the cheap supply or another source and your cabling is acting like an antenna causing the data to become corrupt?

Or Perhaps your F48 is having issues with those specific channels or the actual port. I know you changed receivers but have you changed port on the f48 to see if it was something to do with that port?

 

[Goliath]

Rouge Radio Frequencies, eh? Fitting, I suppose!

 

[scamper]

Check the voltage with a multimeter. More often than not high voltage, around 5.5v will cause a heap of pixels to light up dimmly blue. But it is not unheard of for this sort of thing to happen if the voltage is too far out.
Around 5.2v to 5.3v is good

 

[fasteddy]

Seems like an issue with the PSU that may be causing some electrical interference. The cheap Chinese power supplies are cheap for a reason as they have poor filtering compared to a higher quality Meanwell, So this may be the cause as I have experienced some weird things in the past using these power supplies

 

[LawrenceDriveLights]

Hi guys, thanks for the above advise

Unfortunately it isn't a power supply issue, not sure why it appear to fix the problem before by swapping out the power supply to the F48 pixel controller.

Below is a update video

In summary,
Pixel controller F48 output No. 5 sending data to a differential receiver board that is controlling and powering my tree matrix (3 strings of 168 pixel nodes, 4 stands per sting). Pixels are WS2811 12V from Ray Wu.

I have two of these matrix tress for my display a left and a right (issue with pixels , the first pixels in two of the strings, coming on as red with no data sent to them.

I then connected the left tree receiver board data cable into the output 5, and the issue with the red pixels moved to that tree, but moved over one pixel strand, i.e. the first pixel in the 3rd string was red and the first pixel for the star (also connected to the receiver board come on red).

Now the setups are identical (like a mirror image so speak, so the number if pixels etc on each of the differential receiver boards are the same when I swap over the F48 outputs for these trees).

Yes this is a bit tricky to explain and thus why I did a video, hope it doesn't confuse you to much trying to follow what is going on.

Any suggestions, could it be a possible fault in the F48 output port No. 5, if so are there any other tests I can do?

I also tried rebooting the Falcon F48 with a factory re-set, but this issue was still there

Thanks in advance and I appreciate any help with this.

 

[AAH]

You've linked wrong vid.should be
View: https://www.youtube.com/watch?v=vTVUBHQRIVQ

 

[AAH]

Can you screenshot the webui settings for the faulty outputs vs the duplicate one that is meant to be the same.

 

[LawrenceDriveLights]

Alan, thanks for fixing my link to the correct youtube video
Here is screen shots of the F48 controller user interface

Note; Originally right tree set to outputs 17, 18, 19 and 20. (output 20 is for the wall tree star) which is the sent data from the F48 outputs 17 to 20 to the differential receiver board.

I made a copy of these are set the right tree to outputs 41,42,43 and 44 (output 44 is for the wall trees star) and the data is sent from the F48 outputs 41 to 44 to the differential receiver board.

I have highlighted the output numbers wit h the red arrows in the screen shots.

View: https://www.dropbox.com/s/ssxf1dzvjqyz318/F48%20Red%20Pixel%20issue%20on%20right%20wall%20tree%20150618%20page%201.jpg?dl=0

View: https://www.dropbox.com/s/qocdafsg8wg0y7h/F48%20Red%20Pixel%20issue%20on%20right%20wall%20tree%20150618%20page%202.jpg?dl=0


I also moved the 240v power lead that supplies the differential receiver boards away from all the data cables and also tried it on a separate 240v circuit, but still have this weird red pixel issue.

Any thoughts?​

 

[AAH]

What is on the headings for each of the columns?

 

[LawrenceDriveLights]

Here is a screen shot of the top if the page

View: https://www.dropbox.com/s/pma6efjm83mxgty/F48%20Red%20Pixel%20issue%20on%20right%20wall%20tree%20150618%20page%203.jpg?dl=0

 

[AAH]

I can't see any config issues.
Have you tried the inbuilt test mode to see if the problem remains?
If you close down xlights and anything with FPP on it and then cycle the F48 does it still flicker?
If you swap the cable between the F48 and receiver does it stay with the output of the F48?
Final "software" fix is to try to disable all outputs except for dodgy 1 and see if the flicker is still there.
Final test before I run out of thoughts is to send data at the F48 with da_tester from da_share.com.