Testing Pixel Strings

[cdlouke]

I have a few pixel strings that do not light at all on known good ports. Is there any way to test or is the whole string trash. Hate to throw away a whole string if there is a way to salvage some of the pixels to use as spares.

[dibblejr]

If they do not light from any given point the culprit is more than likely the last working node. You can cut it out and splice and more often than not the others will light up. If they do not try the next one in the chain after the one you cut.

JR

[Orville]

10 hours ago, dibblejr said:

If they do not light from any given point the culprit is more than likely the last working node. You can cut it out and splice and more often than not the others will light up. If they do not try the next one in the chain after the one you cut.

JR

Odd, I have had dead RGB lights at the very last bulb of an RGB strand, and all the lights still worked as they should, except for the very last pixel that had failed, even have had strands where multiple pixels throughout the entire strand were dead, including the last 3 bulbs on a strand that were dead, still functioned.  The good pixels all still worked, only the dead ones would not light in a test on a known good controller and working port via the HU, and in a sequence they worked, except for the dead nodes, all stayed off, but the strand still would work as programmed, even with numerous dead nodes throughout it, again, including the very last or even the very 1st one being dead, all others still lit.   If the entire strand didn't light any bulb, if they were C9 types, I disassembled them and kept the globes, bottoms, rings and housing, I just cut and removed the pixel assembly and tossed it.   That way I always had spares available I could use for good C9 pixels that a globe may get broken,  lost, or discolored, or the bottom/housing wouldn't hold together for some reason.  But if the entire strand wouldn't light at all, I considered it toast and junked it, after stripping it down of salvageable components.

[TheDucks]

Thee are 2 kinds of fails for smart RGB

1)the chip: fails to pass the bucket on (all remaining out)

2)A color LED fail (any color out or wrong), the rest of the string is OK

[dibblejr]

7 hours ago, Orville said:

Odd, I have had dead RGB lights at the very last bulb of an RGB strand, and all the lights still worked as they should, except for the very last pixel that had failed, even have had strands where multiple pixels throughout the entire strand were dead, including the last 3 bulbs on a strand that were dead, still functioned.  The good pixels all still worked, only the dead ones would not light in a test on a known good controller and working port via the HU, and in a sequence they worked, except for the dead nodes, all stayed off, but the strand still would work as programmed, even with numerous dead nodes throughout it, again, including the very last or even the very 1st one being dead, all others still lit.   If the entire strand didn't light any bulb, if they were C9 types, I disassembled them and kept the globes, bottoms, rings and housing, I just cut and removed the pixel assembly and tossed it.   That way I always had spares available I could use for good C9 pixels that a globe may get broken,  lost, or discolored, or the bottom/housing wouldn't hold together for some reason.  But if the entire strand wouldn't light at all, I considered it toast and junked it, after stripping it down of salvageable components.

Well if the very last bulb is bad then there are no more to be affected, so that's not odd. As I stated above "typically" there are going to always be other scenarios. 

JR

[k6ccc]

46 minutes ago, TheDucks said:

Thee are 2 kinds of fails for smart RGB

1)the chip: fails to pass the bucket on (all remaining out)

2)A color LED fail (any color out or wrong), the rest of the string is OK

Add a couple.  The chip passes all 1s so the remainder of the string goes 100% full white.

A color LED is stuck on (either one color or full white) - I assume a chip problem, not an LED problem.

 

[Orville]

23 hours ago, TheDucks said:

Thee are 2 kinds of fails for smart RGB

1)the chip: fails to pass the bucket on (all remaining out)

2)A color LED fail (any color out or wrong), the rest of the string is OK

Only had #1 when a lightning strike close by took out 3 RGB Controllers, and 2 strands of RGB pixels.  Cut each one off, starting at the beginning bulb (bulb #1), and worked all the way to bulb 50, every single one of those were dead,  so I just salvaged the useable components from those strings, {connector, globes, housings}, and tossed the RGB modules into the trash.   But I've never had one just blatantly fail with all pixels bad on an RGB strand from usage where all stayed out.  Took an outside force of nature to do that.

As for #2, this one I've had on quite a few different RGB strands, just cut out the "bad" pixels, and then splice in known good ones, or just shortened the strand, depending on how many bad ones were on it, and their location.

[lightzilla]

I had 3 bags of 5v pixel lights (ea bag had 6 strings of Pixel Bullets) I bought on clearance from Lor. I couldn't get them to light up. I bought a new pixie 16 board etc during the mad grab a few years ago. Got it all setup according to lor instructions. Some how I skipped the part where I need to make sure what numbers were on the pixel lights.   I thought the pixie controller just turned them on automatically during testing when I click on all ports.  Nothing happened. Then I realized I needed to make sure I had the right number (2801,2811,2812 ) etc selected. I believe 5v were 2801. Then they worked. (most).

Those annoying pixel lights that stay on folks already mentioned.  However after 2 years of it staying on white (but did it colour change when it was time to do so during the show) didn't come on white again (1 week before Christmas 2021) when it dropped below -25 etc for a week. So I thought it is burnt out now, but no, it worked flawlessly. 

 

[dgrant]

I've repaired numerous problems over the years with pixel strips but its a sure pain to accomplish. Slicing open the silicone tube, peeling back and holding open with various medical tools, severing the strip then having to tin and resolder in a known good section or more. Then resealing up the silicone tube. Worse when I have multiple types of LPD6803's I bought hoping to duplicate CCR's and then when I got them, turned out, some of the LEDs are mounted differently or they are just flat out different, so you have different color orders because there's no set standard for manufacturing. Put that together with controller that don't like mixing string types...lots of wasted money.  Worse, another set of 4 spools of supposedly LPD6803's and not a single one will work. Anyway, yes, they can be repaired but depending on the cost for new, it might be worth it to replace or not.

[k6ccc]

Lord yes!  Strips are a ROYAL PAIN to repair!!!