I just recently started dedoming LEDs, and up until now it has been successful. However, in my BTU Shocker once I upgraded to a 5 amp driver, I've popped three XM-L2's - one of the bond wires goes "poof". The first time it happened instantly when switching the light on. I chalked it up to me not being careful scraping away silicone around the emitter. I got new emitters, dedomed in gas and cleaned up with alcohol, and I did not apply LED Seal this time. Fired it up and "pop" went another one. I chalked this up to being too aggressive using canned air to dry them off. I replaced the bad emitter with one from my first dedome batch, and this time it actually worked for a good 10 minutes. I just flipped it back on this morning and another one popped.
I'm thinking this is all due to a bad dedome technique. Have any of you guys had the same issue ever?
My one last try will be to dedome in gas, lightly rinse with isopropyl, let air dry, and install. But there goes another $15 if it doesn't work (using scratched dome LEDs from IS). I might just leave the domes on from now on.
The first is an XPG2 in a modded zoomable 1x14500 light with a 3 amp driver. The image of the emitter now shows a large dark patch radiating out from one of the bondwires. The dark patch (which produces no light now encompasses probably half the emitter). However, when I look at the emitter in a magnifying glass with the light off, there’s no visible damage. As far as I can tell the die wires both look intact, and there’s no discoloration visible on the phophor.
The second is also an XPG2. Put it in a modded zoomie with FET driver and fresh Samsung 25r 18650. Turned it on at max power. It produced light for less than 5 seconds then went dark. When I looked at it under a magnifying glass, both die wires were sheared through in the center. I assume the high current caused them to melt. It’s likely the star wasn’t properly attached to the pill so there wasn’t enough heatsinking… but I’m surprised that a direct copper star even with no pill attachment would overheat and melt the bondwires so quickly.
I’ve had some other emitters that were destroyed in dedoming due to my incorrect technique. Soaked them in Coleman fuel, but when I went to scrape off the dome it came off in bits, with some of my scraping taking off some phosphor. Now I know to soak longer.
only one. Like two years ago, I dedomed a xml with heat-and-finger-nail technique. It was a perfect dedome. Then I was going stupid and wanted to see if the emitter still works. So I direct-connected the star with a 18650 without heatsink. It was bright for about 5 seconds and then there was a smoke… The end.
The isopropyl is not so very necessary i.m.o.
Also, my experience (have dedomed 3 LEDs) is that you’d better leave it in the car fuel for at least 24 hours.
Then they end up clean enough, after some mild shaking the LED around in the glass of fuel.Only a tiny bit of residue around / under the bond wires.
I read somewhere here someone ‘scalped’ an XM-L(2) with a razor blade (taken from a modern disposable razor head) leaving a shiny cut surface.
The thing is probably, that the dome (encapsulation) provides a bit of cooling for the bond wires.
I am not sure why this happened to you. Maybe the new driver makes too high voltage. I've had a dedomed xml emitter going at around 6 amps for a few years. And a dedomed Xpg2 emitter at about 4,5 amps on a fresh battery and its been going for more than a year in a light that I'd refer to as my go-to light. That one must have 50+ battery changes due to low voltage warning.
You shouldn't be surprised, that's exactly how fuses work. It doesn't matter how well the device handles heat, if it draws too much for the wire it goes poof.
Only one for now but it was my mistake when I dropped it and snapped one of the bond wires. I’ve dropped others but they all work.
I haven’t driven any xml2s over 5A and xpg2 over 4A but I will someday.
I’ve always dunked the leds in gas for a while and splash them with some isopropyl, I handle them pretty roughly and they all hold up.
I'm still wondering if it's that driver. It is almost too much coincidence to be anything else even though I don't know how that driver could be killing LEDs when it's 3S input 3S output and direct drive shouldn't kill those LEDs. A buck driver shouldn't be able to deliver more voltage than the cells can produce, right?
I have probably killed twenty at least. None of them the way you have. I broke a few bondwires before using Gas. Since using gas I haven’t lost one while de-doming.
The rest I have killed by bumping the bond wires, or running too many amps through them. Too many amps has come from SHORTS a bunch of times. Other times it was from running too many amps trying to push my luck or simply not using the best thermal epoxy. I changed my epoxy and now I’m having much better luck running high amps.
All of the high amp shorts would have happened with or without a dome.
I don’t believe I have ever lost a de-domed emitter using gas that wasn’t my own fault in one way or another. I do NOT use any kind of sealer or alcohol. I take them out of the gas and run them under slow running warm water.
I am convinced that as long as the bond wires are fully intact that they are no more likely to Fry than a regular emitter. I have done a couple hundred of them now.
Jonny - I've done lots of de-domed Shockers and other de-dome'ings (if that's a word) all via gas, and never fried a bond wire. I ruined maybe 2-3 LED's, and usually got a low glow, could not confirm any broken bond wires, and every time it was my own damn fault of grounding contact, bad wiring, etc.
For de-doming, I rinse with isop. alcohol, spray bottle squirt it directly on the emitter - no probs. I don't air blast it though, just let it dry on it's own.
For the Shockers, been using the IOS classic driver, think it was 3.5A, recently used the IOS 4.5A and found I had to jumper the resistors -- never had to do that before, weird. On the 4.5A driver, got 6.1A across the switch on 3 fresh LG HE2's, maybe 6.2A on 2 Efest 35A and a SONY VTC5 combo. Of course these IOS drivers are no longer, but equivalent drivers can be bought.
Only burned bond wires on one de-domed emitter - again, my own fault -- got a nice new bench PS, cranked up amps a bit too high, and poof.
I destroyed four XM-Ls after dedoming but only one went ‘poof’ when I turned it on without placing the MCPCB back to the pill. I destroyed the rest by accidentally cutting its wires while trying to get the remaining silicon off the emitter.
Now I’ve learned my lesson well and I soak my LEDs (with little slices in its dome) in gasoline for at least twelve hours so that the silicon is gone when I retrieve them.
With regards to driving my dedomed LEDs, the dedomed XM-L2 T6 1Bs on by K40 and T08 are driven at 6250 mA and 5614 mA respectively and they are doing fine until now. I used the stock 32 mm copper MCPCB for the K40 and a 20 mm SinkPad for the T08. The dedomed XP-G on Noctigon (not XP-G2) of my OEM FTT (originally an XM-L light) is driven at 4297 mA it is doing fine as well. All measurements were taken at the LED (DMM in series).
Well fourth LED down. I'm not home, so I haven't been able to test which one blew, but it might be from my batch that I messed up. If it's the one that I just did a fresh dedome on, then I know something is up with the driver (although I don't know how). This thing is one hell of a thrower, so I'm not giving up.
It could be that on such a small scale the air does not circulate significantly, and its thermal conductivity is probably much less than that of the plastic dome.
Maybe one could put something on the bond wires.
Never lost an xml/2. Have used various methods. Only thing I think I do differently to what others seem to do is I don’t bother being very picky about cleaning the residue (if there is any) near the wires just leave it be (too lazy).
Now MT-G2s is totally another story can’t do them, not had one real success, (mind you I even destroy unmolested ones) :Sp
