OP updated with mod help questions.
Direct drive from 3 cells? That would require putting a lot of faith in the cells having enough voltage sag to keep it alive.
I have that same driver in a TR-9T6. It was a pain to pull it out too because they just shoved it in burs and all and it basically was locked in place by the retaining ring threads.
I didn't want to soup mine up much as I didn't want to pull the emitters to put them on copper. So I just put a 70N02 FET in that empty FET bay. It increased tail current up to 6 amps. So about 2 amps per emitter (3S3P). Heats up real fast.
Some day I will properly mod the light. Just wanted a modest boost for now and that did the trick for mine.
EDIT: Oh, just using 3 King Kong ICR's in mine.
True, you wouldn’t want to use sony 30amps or heavy gauge wire either. Thin wires and king kongs would work. At least I’m pretty darn sure it would. I would give it a shot but then again I’m not too worried about frying a few XML’s.
I know it would work with noctigons and good thermal epoxy. Might be iffy though if there isn’t good paste under there.
I’ll be using 3 brand new protected KeepPower 5200 mAh 26650s…if that makes a difference.
Overcurrent will melt the bond wires, but too much voltage will zap the dies even at low current. I have done 5 XREs in parallel driven from 2 series 18650s and it lived (and was stupid bright) but only with no-name Chinese laptop pulls... I tried it with INR 15Qs and all 5 LEDs died instantly without even giving the faintest flash beforehand. ZAP!
I haven’t looked up the keepPowers recently but I think they are pretty high amps if I remember right. 4P3S would be the way to go I would think. It’s really goofy though. I have been messing around with these in various combinations trying to get the most out of them and the resistance from the driver itself seems to make a huge amount of difference as well as the wires used. 4p3s should be a ton of amps anyway and I guess even if you did manage to zap it up there higher you would probably lose a lot to thermal sag within ten seconds or so anyway. Try 4p3s and if you are still only getting around 5,000 lumens then you you can always try and bump it up some more by rewiring it later. Also some of these drivers are really easy to get too hot and kill them too. Start without putting the tail cap on for sure. Just ground it with a screwdriver or something and go through the modes that way. If it works and nothing fries and it’s the output you want you should consider potting the driver after that. Also when you do put the tailcap on have your light meter handy and watch for either climbing amps or dropping off really fast. If either one happens turn it off quick before you fry something.
A zener-modded BLF driver would not suffer any of those problems or require a crutch like potting. The switch will have more resistance (and hence waste heat) than the driver.
Is this what you are talking about?
At $16 it is too much. I was hoping for a simple resistor mod or something else that would utilize the same driver. I don’t want to spend a lot on older emitters. Any other ideas?
What is a reasonable US source for these FETs?
There was somebody on ebay that had a partial reel selling them in lots of 10pcs, all gone now and that part# has been superseded, and the replacement is nowhere as good at the low voltages (or rather, the limited headroom) used in these direct drive circuits. The best replacement we've found that's still available new is the AOD510 (http://www.digikey.com/product-detail/en/AOD510/785-1481-1-ND/3603497).
Thanks CC. I think RMM stocks them also. Is this the same?
Yeah, I forgot he grabbed a big horde of them like I did. We cleaned that dude out!
He still has lots of interesting stuff all the time, that's where I get the inductors for the BLF SRK drivers and also the high grade ceramic capacitors that cost a fortune from Digikey. http://stores.ebay.com/RF-Basic-Store
So…just solder it in place and that is all?
Will it be OK w/4P3S?
Sure, why wouldn't it? A FET works just like a relay except with no moving parts.
I'd replace the original FET too, it's unlikely to be as good as the 70N02.
Comfy, the 103 resistor between gate and source, what purpose would that serve?
That's called a pulldown resistor. It just provides a drain for any leftover charge on the gate pin that might be remaining after the PWM signal turns off. Some controllers have internal pulldown resistors so the external one isn't needed.
I haven't replied to your question Krono because I saw you were in good hands. Only thing I can add is a heat gun will be very handy if you want to pull the stock FET. You will want the driver level while you heat it.
I don’t have a heat gun…will it be impossible to use a soldering iron for the FETs?
And I messed around too long and couldn’t get the 70N02s from RMM…and he is sold out…darn.
Ordered some of the other FETs he has…the International Rectifier IRFR3711ZTRPBF N Channel MOSFET - DPAK-2. Will these work?
Hmmm. Soldering a new one on will not be hard at all as you can hit each pad one at a time with your iron. If you want to remove the original one (I didn't bother doing that, but the suggestion sounds like a good idea), that will be difficult with such a large component. Hopefully, someone with experience doing that kind of thing chimes in.
Tractor Supply may still have some of those heat guns on clearance for $14.50. Some times Harbor Freight has their 1500W 2 speed heat gun on sale.
EDIT: Link added.