Random color flickering and sections not lighting

[tmarshal61]

I have several 5V WS2812B strips to mount on my roof line. Yesterday, I mounted two of them with 142 pixels each. Prior to installation, I had tested everything thoroughly and everything worked with no problems. I did not get around to firing them up until today and now they are acting erratically. I have some sections that do not light up or light up randomly. and I have random color flickering. It may pause for a bit then the sequence starts running partially with some more random flickering.


What is the most likely issue? I am running a pretty good distance, maybe 40 feet from the controller. Prior to installation, I tested to see how far I could go without a null pixel and I ran 90 feet with no issues. Still, I decided to go with a null pixel at the 20 ft point just to be safe. I might add that last night and much of the day, it has been raining although, it has not rained in about 8 hours or so.I am running two 12 AWG power wires, I have numerous power injection points, and the pixels light up very bright white with no pink hue, (of course with some pretty major color flickering going on), so I am pretty sure voltage drop is not an issue.


Should I try adding more null pixels or could there be a grounding issue going on. Could it be something else? My gut feeling is that I need more null pixels but, I don't understand why I could get 90 ft with no power injection and now I am having issues at 20 ft. What is the most likely cause of these types of symptoms? It is a real pain to get into the attic crawl space where the wiring is run. I hate to go rip out the the wiring to install more null pixels if that is not the likely cause.


This is my first attempt to put on a display so, I am pretty inexperienced. I appreciate any suggestions.


Thanks

 

[AAH]

What is the pixel chipset (2811, 2801, ink1003 etc), what board are you driving the pixels with and what clock rate do you have the pixels set to run at?
Providing that you have everything configured right the rain may be a clue. If you have managed to get rain inside any joints that can give yo screwy results. Theoretically it will only affect items downstream of where the water has got in so if the pixels look fine to a point and then go screwy after pixel 50 for instance that's where I'd be looking for a problem. The power connections won't be too worried about having water there but if you go water between a power connection and clock and/or data lines it can seriously degrade the signal.

 

[fasteddy]

First wait until you know everything has dried out as this is a known cause of this issue, then if thats not the cause then i would add a dummy pixel, Other members have had similar issues and by adding a dummy pixel this fixed their issues. Its easier to start with a dummy pixel especially with the distances you are talking

The issue is that people use different cables and for the last issue i assisted with it was because 12 guage cable was used for data,, this actually decreased the length of distance that could be done from the controller to the first light and this was only 10 feet by memory as soon as a dummy pixel was added the issue was fixed

So the first and best thing to try is adding a dummy pixel because bench testing and real world install are 2 very different things and a weak signal may be OK on the bench but can cause issues when you have it set up on your house because the cable can act like an antenna picking up interferance and without a strong signal then you can get these issues.

If that fails then I would be looking at ground loop issues.

 

[tmarshal61]

[quote author=AAH]What is the pixel chipset (2811, 2801, ink1003 etc), what board are you driving the pixels with and what clock rate do you have the pixels set to run at?
Providing that you have everything configured right the rain may be a clue. If you have managed to get rain inside any joints that can give yo screwy results. Theoretically it will only affect items downstream of where the water has got in so if the pixels look fine to a point and then go screwy after pixel 50 for instance that's where I'd be looking for a problem. The power connections won't be too worried about having water there but if you go water between a power connection and clock and/or data lines it can seriously degrade the signal.[/quote]

Sorry, for some reason the chipset in the original post was in really small characters. ...edited.

They are the silicon covered 5050 SMD LEDs with the WS2812B chipset. I have a PixLite 16 driving them using WS2811 with the "fast" setting. I tested first in Vixen with update settings running anywhere from 25 to 75. (Is this what you mean by timings?) I also tried in XLights and the d1-E131 Lighting Control Tester. ...same results.
It could be water, I suppose, it kind of acts that way but I am using the silicon covered strips. In a consious effort to prevent water issues, at the power injection points and anywhere I spliced them, I used liquid electrical tape and adhesive lined heat shrink tube over that.

[quote author=Fasteddy]First wait until you know everything has dried out as this is a known cause of this issue, then if thats not the cause then i would add a dummy pixel, Other members have had similar issues and by adding a dummy pixel this fixed their issues. Its easier to start with a dummy pixel especially with the distances you are talking[/quote]

Sounds like a good plan. I will try again tonight and it the issue persists, I will add another null pixel, (or two).

[quote author=Fasteddy]The issue is that people use different cables and for the last issue i assisted with it was because 12 guage cable was used for data,, this actually decreased the length of distance that could be done from the controller to the first light and this was only 10 feet by memory as soon as a dummy pixel was added the issue was fixed[/quote]

"12 Gauge cable" actually decreased the distance? Wow, I would never have thought that. Only getting ten feet also surprises me. This is great information though. I actually measured and cut my cable before I did the tests so that I could use the exact same wires and connections in test that I would be using in the real setup.

[quote author=Fasteddy]So the first and best thing to try is adding a dummy pixel because bench testing and real world install are 2 very different things and a weak signal may be OK on the bench but can cause issues when you have it set up on your house because the cable can act like an antenna picking up interferance and without a strong signal then you can get these issues.[/quote]

I am running the wires through my attic across AC power wires and the like. That could also be inducing noise into the data line.

[quote author=Fasteddy]If that fails then I would be looking at ground loop issues.[/quote]

What is the best way to go about checking for grounding loop issues if it comes to that?

 

[tmarshal61]

BTW, I read about the issues ShellNZ and others had due to the WS2811 vs WS2812B timing differences and the changes Advatek made to the timings in the firmware for the PixLite 16 so, I am running the latest 1.4.0 firmware.

 

[fasteddy]

You should be able to see moisture in your strip if that is the cause. I had a strip in my tree that wasnt very well sealed at the top and when it rained it would start to flicker, also check your plugs dont have signs of mositure

 

[tmarshal61]

The random flickering issues went away by themselves so, it looks like it was due to something getting wet. Thanks for all of your help!


Tom