One question about the DRY…is direct-drive 4-5A @ 12.6V, or is it 4-5A @ 4.2v?
The LEDs are wired in series so the output voltage should be around 12.6V.
So you’re saying that in direct-drive, the emitters will be getting almost fifty watts of power?
Hmm… I think 12.6V is on paper, in real world at such high current draw the batteries would have voltage sag, and the amount of current draw is also depends on batteries quality and all the contact areas. 50W is a bit difficult due to the sag, but if everything has been done well I don’t see why not getting around 40W of power in DD mode.
This is why I warned him about upgrading wires. You can fry emitters in direct drive if you run the right batteries and wires. I have done it. This is why on my mods I only use direct copper emitters like the noctigons they can handle it while the aluminum star emitters would die a quick and painful death. In my J19mod I’m using three 50a sony 26650’s and 20 gauge wire. That would kill an emitter fast if it doesn’t have the best thermal pathway possible. But with thin wires you get more resistance and with lesser amp cells you get voltage sag so that you don’t kill them but you still push them very hard.
No way I’mma flow three emitters and then try to get them all lined-up with the reflector. Honestly it’s more of a “hey y’all, watch this” mode than actually used I mean, how many people buy a TK-75 and have a regular need for Turbo? I was actually thinking about just selling this light as a host and buying a real 3T6 from Fasttech ($8 more than the Amazon knockoff). It sounds like the real one is both miles more powerful than this was when I first got it, and a constant-wattage light as well.
It’s worth it to put the DRY driver in. It’s an easy mod don’t swap anything else and you will be fine. You will have a better light and no blinkies.
Yes the Noctigon is just as good as the Sinkpad for thermal conduction and they are easy and relatively cheap to be ordered. Noctigon/Sinkpad (or any other direct-thermal copper MCPCB you can find) are critical for overdriven LED.
The DRY arrived! (13 days total, BTW)
Now…how to mount it? :zipper_mouth_face:
How much heat is the driver going to generate by itself? Because I’m thinking it might be worth it to remove all the stuff off the original driver to use as a contact board and simply stand it off. Then the driver will look like a 3-tier wedding cake. 
Oh shoot! I was going to order those brass rings with my last Fasttech order and forgot! I wonder if the hardware store could help me at all. Also, these wires on this driver are really thin, will it really hold the current even high, let alone direct-drive is going to put out?
If you can get away with using a thicker wire then do it. It’s always going to get you better effeciency/current, how much depends on a lot of factors but if it’s doable then no reason not to go for it. Driver to pcb wiring should be as thick as you can get away with, especially with high current parallel setups but even with series currents you should see some benefit.
Resistance losses can be surprisingly high on poor quality, thin wires. I usually regret not doing it and end up tearing the light apart again in the future to upgrade the wires. 
Edit: You shouldn’t look at it in terms of what the cable can ultimately handle in terms of current, ie at what point it will burn up or melt or whatever. But rather in terms of resistance losses which will always be far worse on a thinner cable than a thicker one of that same material. No sense in wasting power to warm up a cable!
I haven’t reported back on my use of the brass ring… well, it worked but it was difficult and ugly. I am going to desolder it and use some split copper tube to form a shelf underneath the driver. If I were you, I would get some copper.
I was advised earlier to leave some resistance as these emitters are in series and 4-5A on direct drive might be unwise to out through aluminum-sunk emitters.
Actually…You know how some people end-up using Nanjg 17mm drivers in 20mm pills by soldering the driver inside the host pill’s driver retainer ring? It looks about doable for this light. The only potential caveat/bugaboo is that this retention ring is a press-fit and it would severely hamper removal of the driver in future.
[EDIT] How about this…I go ahead and bridge-solder the DRY driver into the retention ring. Before I put it back in however, I use a dremel on the pill to carve a notch in the bottom such that when the retention ring and driver are pressed flush into the pill, I can still insert a small screwdriver from the outside and pry it out?
What is your retaining ring made out of? Mine’s Aluminum. If you can solder to it, let me know.
With lots of flux and a 180 watt soldering gun I can KIND of make solder stick to Aluminum, but it isn’t that strong and always ends up being a big blob.
I was afraid of that.
[EDIT]: RMM, 1” OD copper pipe is $18 a foot at my local hardware store. Could I buy…3” of whatever you end-up using from you? My idea is to use 3/4” length of it split lengthwise, solder the driver in one end. The copper then would theoretically provide a snug slip fit inside the bore of the pill and also be a great thermal path away from the driver. One could even put aluminum sinks on the chips and fujik/silver-paste them to the walls of the copper tube.
Also…isn’t this driver a bit…fragile, to be capable of a theoretical 40W to the emitter?
I found something to use. A $2.50 copper pipe coupler, slotted and (poorly) bent in to fit.
The thickness is about perfect to be a spacer, but the section I clipped from the coupler didn’t bend in uniformly. What I’m going to try to do is cut it into segments and solder four segments around the radius, hopefully providing more than enough friction and contact to properly sink the driver. Unless someone has a suggestion for bending that piece of metal into a smoother circle. Maybe halve the depth of the skirt? Keeping the copper level with the bottom of the pill, it keeps half of the skirt in contact with the pill. It might be more cooperative then.
I have a chem test tomorrow, and a project for english on Friday as well, so I won’t be trying this until Friday afternoon
If you can find something round around the same diameter as the inside of your pill, with a piece of soft wood gently tap you copper to the shape of the round piece. Hopefully you will end up with the perfect size.
Best wishes on your exams. When you get back to this there is a easy way to make the copper softer and more pliable. Just heat it with a torch until it is glowing red (should just take a matter of seconds). Have a bucket of water near. When red hot, drop the part into the water so that it rapidly cools. It will be much easier to work with, but will stiffen back up with manipulation.
Thanks for the kharma.
I thought that only iron/steel could be hardened/annealed. That should make things much easier, as long as it doesn’t affect the thermal conductivity.
Use a deep well socket clamped in the vise as a form, copper that thin doesn't need any annealing to work with hand tools.
…So I misjudged how tired I was this evening, and it wasn’t as cold in the garage as I thought…
I tried the socket thing. And it wouldn’t cooperate. I tried the annealing…holy crap, I’m sure I cooked all the carbon out of this thing. The copper now feels like it’s made of lead, it’s soooo soft. And just about perfect. It’s almost perfectly round, I’m going to try and solder it in tonight and see if I can get it to fit.