Falcon Christmas

The Lab (Technical Talk) => Designer Troubleshooting and Help => Topic started by: khanklatt on October 24, 2020, 08:11:35 PM

Title: Diagnosis help for improper lighting of my strands
Post by: khanklatt on October 24, 2020, 08:11:35 PM
Since I changed my output resistors, the flickering problem with my lights has disappeared, but I have a few issues remaining I could use some help with.

Most significantly, one of my strands seems to be getting all three key signals (power, ground, signal), but shows random garbage pixel output. You can see this problem at the last couple seconds of both videos below.

Another problem, visible about 7 seconds into the RGBW test pattern recording is that one strand remains on blue. It's also the strand with the misbehaving initial pixel, so I think I have a theory as to something to try there.

The last two problems may also be related, in that the test pattern of the F16v3's "red ramp" (see red ramp video below) is being taken by at least two of my strands as green. Using xlights, I assume this is easy enough to remedy, but perhaps I missed something when configuring my F16v3 universes. 

Any tips of what to take a look at, in particular for the first issue would be greatly appreciated!

Title: Re: Diagnosis help for improper lighting of my strands
Post by: AAH on October 24, 2020, 11:45:27 PM
How is the problem section wired? Random flickering can sometimes be caused by bundling 2 different runs of cables adjacent to each other. It can also be as a result of having a broken ground connection with a power injection ground downstream providing a ground connection but the current path does nasty things and causes gibberish.
Do you have any control over that section of lights at all and if you do and you have a static colour on it and do stuff with the other runs does anything occur.
The red ramp may not be an issue at all. It could be an rgb order issue but also may be a function of how the test mode works relative to channel numbers rather than assigned channels to props.
Title: Re: Diagnosis help for improper lighting of my strands
Post by: khanklatt on October 25, 2020, 01:33:23 PM
Hi @AAH, good questions.

It's wired as a straight serial line: controller -> 15' of 4 conductor 18AWG wire ->  LED string (5m) -> LED string (~1m).

Physically, it's ready for power injection: the LEDs are adhesively mounted to flat speaker wire in the aluminum channel. Electrically, however, I have not discovered that power injection will be required for these strands, so they are there passively and not connected at the end or middle of the strand. If mounting them on an inactive insulated copper conduit would cause interference or something, that is a possibility.

The strand is a WS2815 like the others; it is wired like the others, to three out of the four pins on the output port, one to ground, one to +12V and one to the signal pin. The WS2815 has two signal wires (it has 4 leads, +V, Gnd, S1, S2 where S1 and S2 are redundant), and switching the output signal from one to the other has not made a difference.

Comparatively, of all my strands, this one has the shortest lead wire, and is one of the two shortest (fewest pixels) out of the 5 strands, all of which are wired similarly (down to having the speaker wire in the channel in case I needed to do power injection).

I do have control over the section. When I apply the test patterns, this strand behaves as you see in the video, while the other four, for the most part, behave normally.

Perhaps what I have is one (or more) damaged pixels that is throwing off the cascade of the remaining signals. Maybe I will configure an xlights test pattern of an "ant crawl" (RGB through all pixels) to see which pixels aren't happy. If this pattern is quiescent electrically (e.g. I have only 1-3 pixels on at a time) it might highlight which pixels are faulty outside of strand-related issues (like power draw, etc.)?

Thanks again for the reply, you've given me a few things to think about and try.
Powered by SMFPacks Menu Editor Mod