Another idea for the Nichia LED- is to have a smt wire pad (like LED+) right next to the LED and run wire pad to pad over to the FET. Was there smoke before the blowout? or was it instantaneous?
I had 3 Nichia's and a BLF member donated the 4th, so they could very well be from different tapes. The problem with that theory is that the one that blew up was one of my 3 lol. I might test each emitter before I put them on the revised board.
To be honest I didn't even notice! I was pulling it apart to adjust the XPE led and noticed an odd smell. Not a burning smell though. It was weird. It was at that point that I noticed the LED had blown up. Can't believe I never saw the brown optic lol. So if it did pop whilst being used (obviously) there was no smoke and I never heard or noticed anything at the time.
Looking at the pic, the blown track, on the 4 x Nichia, is the shortest one with the least resistance. It would probably pay to make the total LED- tracks/wires the same. This could be done by moving the micro Vdd track over to the left so that there is two parallel tracks/wires of equal length for the top two LED~~, and bottom two LED~~.
What is the power source, or how many and what type of batteries are being used?
Good idea on the track length. I might try and figure out a way to make the track a complete ring and pull VDD from elsewhere.
Its in an SRK so I'm using 4 parallel 18650 cells. They are IMR though which won't help.
IMR, DD, Nichia 219, doesn’t sound like a formula for success? You and someone earlier mentioned the quad copper mcpcb, detune with higher rdson FET? 
.
Good point. With 219a I think it would be fine - the impossibly high Vf would keep things under control. The 219b has a low enough vF to let you get carried away. Something like 5A @ 3.7v.
I might include a current limiting resistor for each emitter to dial back the maximum drive current. I really don't want to be mucking about with noctigons or whatever the quad version is called.
Why not switch from IMR to ICR? IMR isn’t doing anything useful for you here, just reducing runtime. LG D1 looks like a good candidate here, then if you want to get frisky you can charge them to 4.3v or 4.35v. Since you’re DD with those particular emitters I think the discharge curve matches the emitters extremely well!
Got a link to those particular cells? I'm not across my batteries very much. I'm going to be buying 26650 cells soon so if I can grab them all in one hit that would be helpful.
Never mind found them on fasttech
I have two of these on order from Fasttech. Just a warning that they will more than likely shipped on a ship not airfreight as my order said. I'm not expecting them any time soon.
Yeah I'm probably going to get mine from illumn with my 26650 cells. More expensive but I trust Craig and Co. more than Fasttech.
No cells though, but there was a knife…
Sounds good. Pending the opinions of others, I’d say you probably don’t need to change the circuit if you’re going to use the LG D1 cells. They really look like a perfect match for what you are doing!
I thought you’d be interested to see this (post #83). Comfychair hooked up a 219 on an IMR cell just now - it turned blue. So that’s a no-no.
And mine was on a copper board, you're going to need more than just adding more vias. Drill & insert copper wire will be about the best you can get, though still not ideal.
What's on the backside of the Medusa board? Lots of copper? Could you do this trick with it?
Solder it to the copper sheet first, THEN drill the holes! Otherwise it's a serious pain in the ass.
you always do amazing stuff Matt!
Ouch! If this keeps up, someone will earn the nickname of “smokey”. I appreciate the limits to which the BLF community will do for the sake of science.
It survived, it still does the same output on an ICR cell as it did before I hit it with the 25R hammer. I did swap it out for a XPG2 S2 1A though.