I wasn’t sure where to post this, as there isn’t an automotive sub forum. This doesn’t really qualify as an LED bulb… but the topic is more about the quality of LED mounting to a PCB, which feels more “DIY”?
Anyway, about 6 years ago I was having an issue with my Audi A3 license plate lights. The incandescent bulb did not have proper ventilation (bad design of the housing) so in time it began to melt the housing and eventually short out the circuit. After I had the whole license plate light unit replaced, I made sure to use an LED substitute. Thankfully at that time, there were already a couple makers providing these specifically for my car and others.
All in all they worked fine up until just recently when one of them went “out”. I was puzzled as to what could have caused it.
Viewing as-is, you can see the “cluster” of emitters behind the lens.
Then, taken apart… all is revealed:
The unit is a bit heavy… which is that slab of metal inside that I figured was for thermal management on those 18 LED’s blazing all at once. You can see it in this last side view photo. But then you see the gap and realize… it couldn’t possibly be serving for thermal management. I’m assuming that the electrical load spread out across 18 LED’s helps minimize the heat build-up.
My old car had that with the hotwire bulb melting the housing over time.
My newest one, I substituted NW 194s/921s in the lowest lumenage I could get away with. I don’t like when it’s all lit up like a spotlight, just a nice glow. So I bought 6-8 pairs of NW LEDs from FT and picked whichever ones I eventually settled on.
Ie, there’s no need for 18 5050s to melt their way off the pcb. If anything, I’d stick a series resistor to limit wattage/current to a safer value.
Other marker lights, I want as bright as possible, just not the tag lights.
Looks like water was getting in there.Here’s something Google found-
“As stagnant water contacts copper pipe and lead solder (simulated soldered joints), a corrosion cell is formed between the metals in solder (Pb, Sn) and the copper. If the resulting galvanic current exceeds about 2 μA/cm2, a highly corrosive microenvironment can form at the solder surface, with pH < 2.5 and chloride concentrations at least 11 times higher than bulk water levels. Waters with relatively high chloride tend to sustain high galvanic currents, preventing passivation of the solder surface, and contributing to lead contamination of potable water supplies. The total mass of lead corroded was consistent with predictions based on the galvanic current, and lead leaching to water was correlated with galvanic current. If the concentration of sulfate in the water increased relative to chloride, galvanic currents and associated lead contamination could be greatly reduced, and solder surfaces were readily passivated.”
My guess is that the LEDs got too hot, melted the solder, and fell off the board. Have you compared to the left side to see if there’s any difference in thermal management?
Yeah, they are a bit brighter than I really want, but it’s not terribly so. It’s about on par with others I’ve seen OEM on brands like Lexus. I don’t want to mess with a resistor, because it’ll likely trip the bulb-out warning on the dash.
I think you’re right. Even though the gasket was still intact, perhaps there was a slight ingress somehow.
I would definitely not try to resolder these LED’s back into place. The new set I got was only $13, which means $6.50 a piece. Definitely not worth my time trying to fix the old one. But also, considering it lasted 6 years is pretty good.
But, this is over a 6 year period… and based on what I’m seeing, it appears like some metal component of the solder rusted. So it could be that there was moisture inside it. The left one seems fine with all LED’s intact. I didn’t open it up though. The new ones mounted well and with the thicker gasket I’m expecting they’ll be fine. My car is just about 12 years old at this point and I don’t see myself still driving it even 5 years from now. By then, it’ll be time to move over to electric.
