The last thing is the cost. Add the price of AA for 40 or 50 units? No thanks.
I agree. I was just back in the shop and had some of the 4x drivers and thought, you know? I should just do the 1.4amp drivers, because of the flat surface and ease of doing it, but I would still use XM-L most likely. You think XM-L at 1.4 amps would produce more lumens than a Nichia? Nichias drive the price up too.
From the JB Weld FAQ
Nichias in everything, unless the XM-L is a T6 4C.
I thought you had computer problems. Isn’t it past your bedtime?
LOL, I am olny razzing you.
Probably was my method or something. I have burnt up drivers with JB Weld, that's why I went to AA. Plus JB Weld takes a lot longer to fully set. Can you tell I really hate JB Weld?
The biggest reason for using XM-L? It's because the TIR optics are made for XM-L and they are too big in diameter at the base, to fit a Nichia well. With an XM-L, they fit tight around it and "push on", but with the Nichia's smaller size, they are sloppy and I have ruined several Nichias because of the play.
I really do like the Nichias better because of high CRI, but I also like an XM-L on high mode, just for the wow factor when others see it turn on.
Maybe I should run a poll on it and see what the majority says.
I could just isolate the whole pill by painting it with black paint. That would isolate it from the head and still use the current method for the driver. Between the paint and some thermal compound (I got probably .003" clearance between the pill and body), it should isolate it.
Yes? No?
I’d say no… paint has a nasty habit of not staying in place when you want it to. Scratches through too easily, etc.
I suggested Nichia for the 1.4A driver that would go with it and I know you would enjoy testing them more. 8) The 10 degree optic fits snug on the XML chips so no squirming there if you go with XML. I’m sure you will get endless requests for every tint(3D is right on the black body radiator line 0:) )and drive current possible and maybe you could offer one or two choices. At least the hosts would be identical. If we could figure a way to use the Nichia, this would be a done deal to me. At 1.4A an XML or xp-g2 will be brighter than a Nichia.
If you pot the driver and gently press it into the pill so as not to press the annular ring against the copper coupler it would work. You have to make sure your “step” doesn’t contact the ground tabs on the chips either. Anything that can stand the heat, maintain the separation, and be a permanent adhesive would fit the bill.
Who makes the holder for that XML optic and do they make one for the xp footprint?
Have no idea of the Mfg. It’s a Chinese copy cat TIR. They make the 15, 30 & 45 degree on this style and I am sure they make others, but I have not seen any XP-G in this particular style. The other styles I have tried, I don’t care for. I like this particular knock off better than Carlco.
I'm having a bad day I guess. The new pill design will not work. Read post #8 for the info.
I made a decision with this build, that if I could not make a simple, bulletproof and proper pill with hand tools, that I would not do it. I am at that point right now. I know a lathe could solve it all, but then it would just be another GD CPF light. I will not glue this thing together, like I have in the past. It's not good, not right, not worthy and I just won't do it any more. Gluing is BS and it's expensive too. I'm tired of *@#$%*! rigging every mod and not making something dependable. These mods I do are as far away from EDC as S**T is to Shinola. Not happening any more. Yup, it's a bad day.
The only way to complete the setup "as is" would be to just use a one sided driver (NANJG 101Ak 1.4A), where the battery side was smooth and the contacts were used as is. The outer ring of a 101AK would work and I could even add a Nickel Silver washer on it, to make it last longer. I worry about the friction of on/off, on/off so many times. Anyhow, that brings it back to a Nichia and an overall weak beamed light. I just don't see the sense of doing a 1.4A driver. There's a S***load of 2AA 1.4A lights out there already. It's just not unique that way... I'm not about to add chips to it either. I can't solder that small stuff. It would take years to complete that, LOL.
Maybe tomorrow will be better.
sorry about the bad day O-L! have a drink and go think about something else for a while 
OL, it sounds like the heat is causing problems. Why not try out a bit of wire glue? No heat at all is required.
Maybe because wire glue sucks… doesn’t handle current well. Tends to not stick well.
Well, so far, the majority vote on the emitter favors the Nichia and that would solve the soldering problem, as the driver would not have to be soldered to the coupling and it takes care of the isolation problem too, so if the threads ever were to bare, it would not cause permanent on. I just don't like the use of the Nichia. I'm a bit tired of them. He-he, I'm a bit tired of all of them actually. Once you've seen two, you've seen 'em all kind of thing.
In the long run, the 1.4A driver makes the most sense and I should realize the limitations that using hand tools inflicts. I'm just sort of stubborn and don't cotton to No. Plus, I don't want to reflow 25 or more leds onto XP-G stars.
Hey, maybe I should have put XP-G2 in the poll too... Never thought of that. It's not as nice a beam as the Nichia (to me), but it's popular right now.
IS has XP-G2 in a 3D tint.
I offer my sincere appologies if I damped your enthusiasm for this project. I would like you to do it and for it to be successful. FWIW, I think the new pill design is a good one and potting the driver into it with Fujik instead of soldering should be very robust. An additional nickel/silver disc is not needed if the driver ring is tinned(I haven’t detected any wear there yet).A short section of copper pipe even with the step would support the chips and keep the annular ring and ground tabs above the step. I If you wish I could supply the drivers with chips added if that would help. I would enjoy making a contribution. Here is a pic of the XML pill I made this way. Top left.
Uhhh, I don't think you could dampen it. If anyone else, besides me, were to say, it wouldn't work, I would have just proved them wrong. You just gave options, just like the others here did, nothing wrong with that at all.
I see that the design I had, has a flaw, just like the others. The flaw is simply, multiple solder joints on the same copper parts. Every time I have a different solder joint in more than one different step, where it means a previous solder joint will be exposed to heat, it's a nightmare. It's always been a nightmare from the very beginning.
It stems from me not being able to make threaded parts, or make parts that fit right with a lathe. So, it's my own fault that this stuff comes out the way it does, no one else's.
As far as gluing, nope, not going to happen any more. Every driver that I have glued together, in these AA lights, has failed, every one... The glue has broken free from the heat or the poor ground has failed. Now, with the Nanjg 101-AK-A1, I can see using it in the pill, because, as you say, it would be supported well and probably not fail. I think all in all, everything points to using a single sided driver, unless I was able to make a machined insert for the double sided driver, where it could be insulated and have a solid contact plate.
I understand the way you did the driver, with the wire through the holes (what's it about 32 ga wire or smaller?), and through the top plate too, but I would not even attempt that. Too delicate and fine for me to try and way too much work for 25 or more drivers.
I thought about tinning the driver ring, but solder is also very soft. Will it still hold up under a thousand friction turns? I am not sure, but I think it won't. I figure the Nickel Silver Disc on top of it would, that's all.
I will tell you the same thing that I tell others. I do sincerely appreciate the offer of making something for me, but if someone else does it for me, it its theirs, not mine. Absolutely No offense meant by that either, just the way I am.
I’ll try to think of solutions.
Anyway, I know exactly where you’re coming from about building things so they continue functioning as intended down the road without maintenance (what I think of as ‘user intervention’). For example, when I pull a flashlight, gadget, or tool out of a cabinet one month or even five years from now, I wanna be able to expect that it works just like it did when I put it away without having to fuss with it. In my eyes, that epitomizes perfect function.