WS2811 Voltage Drop Issues

[tmarshal61]

Hi,


I am trying to build some landscaping path lights using RGB pixels. I bought some low voltage landscape path lights from my local hardware store and replaced the lights with 12V WS2811 RGB modules. I am running 12 AWG low voltage landscape wiring injecting power to each light along the way. There 31 lights spaced about 2.5ft apart.


I wired up all the lights and laid them out and tested them and everything worked like a charm. All of the lights worked all the way to the end. Then I buried the cabled and that is when the trouble began. About the first 9 pixels light up fine then I start to get some flicker. The 12th light stays on solid red, and none of the other lights light up after that.


I have dug most of the lights up and checked the voltages along the way. What I am seeing perplexes me. The 12V positive supply seems fine all along the way. The data line starts out at about 5V out of the controller and drops a bit through each light. By the time I get to about the tenth light, I am at about 3.5 V on the data line. That is about the point where the problems become visibly apparent.


What could be causing my issues? Isn't the data supposed to regenerate after each pixel. If this is the case, I would expect the voltage to be boosted back to 5V coming out of each light but this is not what is happening. Am I just misunderstanding how this is supposed to work or is there something fishy going on? Anyone have any ideas?


Thanks,


Tom

 

[David_AVD]

Is the cable double insulated, especially for the data wire?

If it only failed when the cable was buried, maybe it's due to the capacitive effects of the (damp) soil around the cable.

 

[tmarshal61]

No but, I buried it in waterproof PVC electrical conduit. Only two or three inches are exposed between each electrical box and light.

 

[David_AVD]

Does it all start working again with the cable free from the ground?

Did you do anything else to the wiring after you did the initial testing?

 

[tmarshal61]

I just ran a 12AWG data line from the controller to the point past where the issues were occurring and once again, it lights up about 11-12 lights. I then moved the data line down another 12 and I still get the same thing. Only about 12 lights light up.

 

[AAH]

Based on the maths you shouldn't be having any problems. Measuring the data line doesn't necessarily tell you anything as the voltage of the data line is dependant on what data is present at the time and as part of the data is stripped off by each pixel the voltage wil actually change at each pixel. Providing you have voltage at each pixel which you appear to have then data integrity and the actual data are about the only things that will give you problems.
What are you driving the pixels with because it almost sounds like you are driving 31 channels rather than 31 pixels (93 channels)?

 

[tmarshal61]

What are you driving the pixels with because it almost sounds like you are driving 31 channels rather than 31 pixels (93 channels)?

You are exactly right. I was so sure I had it configured right, I just went straight to thinking it was an electrical issue. I was originally testing the lights with Vixen. (configured correctly), but after I buried them, I decided to used XLights, and in my hast, I selected 31 channels, not 93.

Don't I feel stupid.

I sure did a lot of extra work for nothing.

Thanks for the assistance.

 

[David_AVD]

When the result changes you really need to ask yourself what other conditions changed. I assumed (silly thing to do) that you had the same input(s) for the test and final installation.

Gold star to Alan for seeing beyond what was written.

 

[tmarshal61]

There may have been an electrical issue too. It was the flickering lights that sent me in that direction. but I guess messing with all the connections cleared that part up.


Thanks for your help guys.

 

[mborg10]

I had similar experience with my lights last year. Turns out that I did the same thing. 31 channels instead of 93. I also had a pixel staying red so I changed the pixel just before it and it worked fine. DUD pixel.

 

[nzlongfellow]

Don't feel bad. Same for me the other day. tested one way, all fine. Changed program and any half ran. result---- not enough programmed outputs --- and a DUD pixel.
I think the heat gun may have fried the last resistor of the last pixel when the shrink wrap was applied.