Well good news and bad news. It's no longer gold, but somehow the driver ended up dead. Or, one-third of the driver is dead, the section for LED3. I removed all the junk I'd added to it, cleaned everything, looked it over a hundred times with magnification and good lighting, I don't see anything. I guess I will use this driver as a contact plate for the MT-G2 version, and the driver from the other light in this one.
The aluminum plate in the first pic is 1/4" thick, with a step cut on the bottom to set the internal spacing. When the bezel is screwed down tight, it's right at .005" shy of being fully seated against the head, has to be that way as the stars and reflector just sit there loose until the lens clamps it all together for good thermal contact.
I still need to machine off the lanyard loops that got all bent up in the War Against Loctite. Final finish will be just buffed with wet green scotchbrite, as I hate polished aluminum. Like this, if (no, when) it gets scratched or banged up, I just hit it with the scotchbrite again and it's like it was never there. Plus it looks a bit like titanium and titanium is just sexy.
given how finickity the MT-G2s are proving with the smaller reflectors (no huge surprise really, considering it’s size), why not mix’n’match? You could wire 2 XM-Ls in series and then in parallel with one MT-G2 - if you can pump in 5A (I know, I know), each string would get 2.5A, which would be a ton of light and relatively efficient. Not exactly a simple option, but might be worth thinking about.
Abso-freaking-lutely amazing job you did with the miters in the reflector array, Comfy!! 8)
I've often thought of doing what you did but never tried because I would frustrate myself trying to get it as good as you did and would probably not succeed anyway. Kudos!
Damn shame about the chrome coming off but using the heat like you did left it no choice but to jump off the reflector like a scalded cat. If you had a band saw you could make the initial rough cuts pretty safely without destroying the chrome then come in with the hand file to do the fine tuning needed.
Again, great job! Once you have them polished up I can't wait to see what Lux reading you get from the entire upgrades you've made to that light. ;)
Just so you guys don't get the impression I'm some kind of infallible wizard, this is what happens when the angles are wrong. The three are touching at the outer corners with a gap at the center points - the angle between the two cuts is too shallow. It can still be saved, but changing the angle on all six facets (and altering them all by the same amount) after they're all cut this far is really not fun.
LOL, well, I ruined that batch of three. Trying to fix the first mistake made it worse, but in the other direction. What started the chain of fail was that I didn't sit the reflectors in place and sight down the center of the neighboring LED, like I did with the original batch, and like I've done with all the others. I just figured "bah, I'll just make the initial cuts shallow, to leave enough meat to fix it in the fine tuning phase." Wrong! Oh well. It's not like P60 reflectors are an endangered species. I already had more on order.. I think I always have more on order.
Man, this f#&^$%* light. It's out to get me. I fixed the mystery fault in the driver, by removing all 5 of the whatever-they-are soic8 chips, cleaning everything thirty-seven times, and then reflowing them back on. And somehow, after that, all 3 LEDs work now.
But... not exactly. One of the LEDs has developed dead spots at random locations on the die (bottom right in pic below, marked with red sharpie on the dome so I remember which one it is), and I have no idea what happened to it. I didn't mess with the domes at all, it just lost its will to live, I guess. And another of the LEDs has developed a giant gash across the top of the dome, out of nowhere. I have spares, but that's two U2s down the toilet. The one with the scratched dome can be de-domed gasoline dunk style and still be usable, the other one is trash.
And the reflectors still aren't polished. I'm procrastinating. Because I know this light will fight back like it has since I first took it out of the box and found the tailcap was welded on with half a tube of that damned red loctite. Oh, and I still need to mill off the lanyard loops. Which I'm looking forward to even less than polishing reflectors.
Reflectors still not polished, LEDs replaced with XM-L2 T6 1Cs. Waiting on an order to come in, then gonna strip down the original driver and use the stock MCU & switch to run three MCU-less 105Cs.
Minimum OD after machining off the threads looks to be around 33mm without even shortening them at the front face, which will probably have to be done anyway due to the limited space available. Will easily fit in the 52.5mm head, though the LEDs will have to be moved in tighter than is possible with the 20mm stars. So one-off copper pillars it is, then.
Being made for the SST-50, the opening is just the right size for the MT-G2 dome with no cutting!! (phew! that's where I've made the most mistakes up to now, glad to see this problem area eliminated)
But the best part is, the beam pattern with the MT-G2 is fantastic. Very tight center spot, and a 'flat' smallish outer spill (very little change in brightness from the inside to the outside). Very different than any other reflector I've tried with the MT so far. This reflector could even be good enough with this LED to justify a very compact single MT-G2 build, though I have no idea if there are any hosts out there that take this reflector in unmodified form.
Again orsm. There is lots off hosts out there to take this reflector. They just have to be made. I would like to know what this led would be like with smaller currents. Imagine this light l made with your led.
I haven't measured the current or what the voltage ends up at, but it runs pretty nice direct drive from three cr123s. That would be a super compact setup. With only two cr123s it only pulls 650mA, as I'm sure the voltage sags too much to let it eat more current.
No, not a single 14500. That would need a boost driver, and that means no 7135. And I don't think there are any boost drivers that push enough current to make it worth doing. With two 14500s you could use a buck driver, but you'd have limited run time before it drops out of regulation (from what I understand that usually happens at 1-1.5v above the LED's Vf, depending on how well the driver works). Might be interesting to see what happens with 2x14500s in direct drive though. I've never used any of the smaller cells so I don't know what kind of current they're capable of, how much they sag, etc.
