Curious if anyone has seen this. My Lavalier has mostly been sitting around since November 2022 (it’s really cool, but like so many things I bought it on spec that I will someday have a use for it). Once in a while I see it on the shelf and plug it in and admire it for a few minutes, and make sure the battery is charged.
Anyway, today I fired it up and noticed that a section of the “display” was dark. After confirming it wasn’t just a weird pattern, I decided to take a closer look and see if I could spot anything to troubleshoot. I identified the last working LED in the chain, and saw nothing visually obvious like a broken trace or burned component. But after some poking around, literally, I found that pressing down on the first dark LED caused at first some flickering and eventually the whole thing came back to life. Unfortunately I didn’t take any before pictures, and I don’t want to keep messing with it to make it fail again, but hopefully the description is straightforward enough.
This seems like probably just a loose solder joint, or I suppose it could be something in the LED chip/package itself (bond wire?) that responded to some flexing. I’m thinking maybe I should carefully hit it with some hot air to increase the chance of my fix sticking, if it was indeed a bad solder joint.
I may have been mistaken about the loose connection. I reconnected the thing after writing this post and having lunch, and the failure was back. But while I was getting my phone to take a picture, it started working again, and I am unable to gently persuade it to fail again.
There’s a possibility that it had nothing to do with me pushing on the LED, it might just be that there’s one pixel that works intermittently, or that starts working after a minute or two when it warms up. I’m going to leave it alone for an hour and see if I can reproduce this, then I’ll report back.
It’s possible, and if bad we can swap it out for you. If you do want to try hitting it with hot air, you’d want to bake the board to dehydrate the LEDs or it can cause damage.
This has been quite a roller coaster. I was feeling stupid for not even considering the battery, as when I put in a fully charged one, it seemed to be fine. So I reassembled everything and set it aside. When I saw your reply I plugged it back it and there it was again, so I finally caught it (this is with an all hsv(0,0,0.25) program for testing):
I recorded two minutes of video in case I could catch it spontaneously coming good again, but nothing happened, so I disconnected it and got my phone set up on a tripod so I could try poking at it on video. And as soon as I plugged it back in, it was working again, so it made a fool of me again. But while I was playing around with it, I did manage to induce a few (different) glitches by gently flexing the board.
Last one, I promise. I got the failure mode on video this time, and confirmed that pressing on the “bad” LED affects it.
I might try the “repair” but the symptoms seem a bit too random for me to have much confidence that just reflowing that one chip will fix it permanently.
Closure, of a sort: after baking it and confirming it was still faulting, I decided to try the reflow and I gave it what I thought was a gentle preheating with hot air, but as soon as I grabbed the chip with my tweezers, I gouged the plastic. I guess it doesn’t tolerate much heat from above, which makes it tricky to rework. I really thought I went in with a gentle touch but it was enough to take a couple of chunks out:
Funny thing about it, though… it works now. So I’m just going to back away slowly, put it back on the shelf of toys I might play with some day, and take these lessons with me.
