Experiencing and insane failure rate on WS2812B LEDs… Was hoping for maybe some insight. I have custom manufactured panels that I have a contracted manufacturer assemble for me. I have 2 boards, 1 with 109 LEDs, 1 with 159 LEDs. Currently I am experiencing a a panel failure of 65%. The LED failure seems to be around 10%. Each panel I have, has at least 1 dead LED if not up to 5. The LEDs seem to fail after about 100hr of runtime. Mainly run on white. What happens is that sometime after 100hrs (sometimes before) 1 LED will no longer respond, causing all LEDs down stream to no longer respond. When apply pressure, sometimes the LED will kick back on, until pressure is relieved. Sometimes when failure occurs, there’s physical damage to the LED visible, other times there is not. I’m at loss here. Has anyone else experienced such failure rates?
Could be a heat dissipation issue, especially if you are running the LEDs at full white. The fact that some of the LEDs start working again when you press on them points to something weakening your solder joints (bad solder to begin with?) which is often heat.
Someone that I know who works with LEDs professionally (and has been doing so for decades at this point) often rails against the poor heat dissipation that many of the modern led packaging has.
The solder joints seem to be ok, we’ve put the panels under microscopes, and dont see any cracking, the only “fix” we have found is to remove the failed LED and hand solder the new one. We have 6 panels that have been hand soldered, and they dont seem to show any troubles
Back to the original point though - I suspect you’re having heat dissipation issues - I’ve see projects with led failure rates like you’re describing when there’s no ventilation (and/or LEDs are too close together while being run at full white), etc…
@Leo_Bettinelli This is a hardware problem, code doesn’t matter (outside of possibly using FastLED’s power management to minimize total power draw (aka heat generation)
@Decept_Studios I experienced very similar issues see my old post… https://plus.google.com/106626345342202981932/posts/TqWBtiHJHjX
I finally concluded that I was overheating the devices with my hand soldering. My understanding is that people using strips do not see that problem.
How did you solder those devices to begin with ?
Even with a reflowing oven, it could have been set too high or too long for these devices.
I’ve had this problem with strips and rings. The best solution I’ve found is to limit the maximum brightness to 200 or maybe 230 and to fix them with thermally conductive glue.
Hi @Decept_Studios ,
I found that, once you damage a device with heat, it just seems to not tolerate some level of heat dissipation, I guess proportional to the initial damage !
That device driven orange just got hit harder by heat for some reason. I experienced similar behaviour as well !
can you tell me how your devices were soldered ?
The WS integrated LEDs are really sensitive to the reflow process, and small deviations outside of the specs or environmental moisture can cause high failure rates. Were you using fresh, just unsealed rolls of the LEDs? If they have been opened before you may need to bake them at a lower temperature (100C?) for 24hrs to remove the moisture before reflowing them.
My two cents that the joint for the LED and the IC chipset has been degraded due two posible reasons: Wrong reflow oven profile (worth to see the graph) and/or handling which includes the moisture. Cold solder can happen too