Jeroen's Project Journals

This project is a wrap. The LED spots worked like a charm, but the light-intensity was simply insufficient, even with 8 LED’s instead of the original 6. And I can’t fit any more WS2812B’s in the same PCB footprint. I looked around for higher intensity addressable LED’s, but there’s nothing readily available through the usual channels, at least not that I could find. And I would be worried about the heat generated anyway. So I abandoned this idea and replaced my cool DIY RGB LED spots with standard, single color, store-bought spots. Too bad, but it was a fun project anyway.

There was no short in the cable connecting ledspot 3 and 4, so I took spot 4 out of the chain and replaced it with another one. I couldn’t find anything obviously wrong with spot 4, so we’ll see if this does anything. So far it works.

So this is indeed some underlying issue, since the exact same LED (well, a different LED in the same spot) blew its dataline again. I’ve asked what this could be on /r/askelectronics. Most responses were not very helpful and there was some annoyed patronizing, but I think the most promising suggestions were that it may:

1. Simply be a short between data and GND in the cable connecting spot 3 and 4
2. Be an effect of the self-inductance of the wire.

So I looked into the issue that half the spots suddenly stopped working, and eventually found that the data line coming into the first spot in the string that no longer worked was constantly high, instead of usually low with the occasional burst of traffic. Looking further into THAT, I found the last LED in that spot looked suspicious:

We’ve lived in our current house for over 15 years now. In the early years I did quite some home improvement projects, and several of those projects involved installing recessed spotlights. The most affordable options back then were kits from Massive, a brand owned by Philips. These kits included 12V halogen bulbs in the GU4 form-factor.