djulien wrote:huh? Wouldn't it make more sense to click Checkout when she is *not* looking? :unsure:Any guesses as to why one white pixel would affect the upstream pixels? If they are the WS28X1 style, there is serial data being shifted in. I could see that a short caused by water could turn on the r, g, and b LEDs in one pixel, but how would they get to other pixels? (unless it was moisture, and it made its way to the other pixels also). Was the string/strip mounted horizontally, vertically, or (other)?
don
When I had a bad pixel, it either turned the rest down the line white or not on at all. I found it wasn't always the first "bad" pixel that was bad, but the previous pixel.
As far as I know the first pixel gets information for every pixel on that string, it grabs its information and passes the rest down the line. When a pixel is toast, it corrupts the data highway so to speak, the other pixels down the line are no longer getting data or they are getting incorrect data.
That is the only reason I didn't put any pixels on my roofline. If one went out I didn't want to have to replace it 30' in the air..... Grizwold may bounce of the roof, but I don't want to be "that guy" that died putting up Xmas lights.
thebaronn wrote:As far as I know the first pixel gets information for every pixel on that string, it grabs its information and passes the rest down the line. When a pixel is toast, it corrupts the data highway so to speak, the other pixels down the line are no longer getting data or they are getting incorrect data.
Yes, that is my understanding also. I could see how the following pixels would be dark after the bad one (getting no data). I don't understand why they would be white, though - I thought the WS28X1 chips needed serial pulses to set the colors, and for the following pixels to be white would mean that they were getting the pulses (all 24 of them). Now maybe a data line that is shorted to +5V is interpreted as white by the pixels, but the spec sheets say there needs to be a transition in there. If they don't really need the transition, then I guess white on the following pixels makes sense.
don
i use gece for my roofline and have a couple of bad pixels and all the rest still work. But ones from ray sometimes dont but my strings from last year work fine even with a bad pixel or two. Like i mentioned this year not a chance.
I definitely lost a few colors here and there with my 168 GECE's on the roof ridge - not horrific but probably 4-6 with one color missing. I have since taken them down but have not set them back up to troubleshoot yet. I lost none of my TLS3001 pixels from Ray though and about half fell from 20' onto the ground and stayed there for about a full day in solid rain/puddles before I got to them so they got soaked. I put no other sealent on them.
I have only ever lost 1 color with just under 200 GECEs in use for 2 seasons. It actually went between seasons. I have had 2 other minor failures where data wasn't being passed to the rest of the string. Those pixels still received data, just wouldn't forward it so they're in parts bins for either end-of-pixel-string use, or other random side projects.
I had 752 GECEs out this year and lost 15 of them. Two did not pass data (assume bad chip or wire) and the other thirteen lost 1 or 2 colors(assume bad LEDs) . We had high winds and heavy rain this year, but i did not see any sign of water intrusion in the GECE i dissected. The wires and LEDs are potted in an opaque conformal epoxy type coating.
I think the issue why some pixels go all white has to do with water in the back of the pixel. In some configurations of the 2801 pixels, the three LED cathode terminals on the IC are adjacent to the ground wire by the back edge of the pixel. Get some water in there and you can short them out to ground and turn on the LEDs. Who knows what residue is on the pcbs from manufacturing, it may not take much to create a conductive path. It also sounds like that when they dry out the All white pixels often return to working order. I have read on other sites where folks dissected the pixels that did not turn off after drying and scrubbed the pcb with solvent and the pixels started working again.
The issue is likely the various techniques that the pixels were made. I think the silicone molded pixels are likely worse since the silicone may not adhere to the wires as well as the newer epoxy encapsulated pixels. It seems like folks had a lot of issues with water on the silicone molded units that had the wires come in the sides of them. Hopefully the newer style pixels with better sealing will solve this manufacturing issue.
Most pixels were not designed for the use we are using them for, they are designed to go inside of a sign enclosure with just the led end in the weather. I initially went with the GECE since they were designed to be outdoor lights, but given the failure rate I saw, I am not sure it made much of a difference.
budude wrote: I lost none of my TLS3001 pixels from Ray though and about half fell from 20' onto the ground and stayed there for about a full day in solid rain/puddles before I got to them so they got soaked. I put no other sealent on them.
Hey Brian,
Which 'flavor' TLS3001 did you get from Ray...the square ones or the round type? I've been looking at Ray's TLS3001's and was curious about the water aspect....
Santacarl wrote:Hey Brian,Which 'flavor' TLS3001 did you get from Ray...the square ones or the round type? I've been looking at Ray's TLS3001's and was curious about the water aspect....
The round type - I do have them secured to wire rope so there is no strain on them.
budude wrote:The round type - I do have them secured to wire rope so there is no strain on them.
Sorry, not trying to hijack, but what gauge of wire did you use and how did you attach the pixels?...with cable ties ? did the pixels move ?
budude wrote:The round type - I do have them secured to wire rope so there is no strain on them.
Thanks Brian....That's interesting and encouraging since the round ones are the lower IP rating at 66 while the square ones are rated at 68.
Have you had them long enough to see any UV impact on the wires yet? I've heard that some of the pixels, since most were designed for inside/enclosed environments instead of outside environments like we use them in, don't fare well under repeated sunlight exposure..... I live in the south and get LOTS of sunshine that time of year.....was wanting to do my homework on that aspect of pixels too....but haven't seen a whole lot of data on it as of yet....
