In JonnyC's original source code it says:
#define SWITCH_PIN PB3 // what pin the switch is connected to, which is Star 4
...and PB3 is pin #2 on the MCU.
If it only works with the switch grounding pin#3, and not pin#2, then somebody altered the pin assignments in the firmware.
What cells are you using? It will only do the big numbers with something like INR 20R/25R or Sony VTC4/5s. On tame ICR cells it isn't going to pull more than 10-11 amps.
Well, good news and bad news. That 3000 lumens was on 3 LEDs. Apparently one of the defined emitters was OK for 2.8 amps but it blew with the new driver. The bad news is I don't have any more XM-L2 U2 emitters to reflow and dedome. I need to order some from Illumn. At least i know it works now and will know what to do in the future with my other SRK driver. :D
Comfy, I will break out the good camera and get a macro of the second driver for you. Hopefully it will show what you wanted to see.
Oh, and thanks again, everyone, for all your input and advice. You guys rock!
For the record, I was using VTC5 cells from Illumn. It actually is brighter on my Panny PD cells. I hadn’t done any resistance mods on the tail springs yet but I just finished adding 20ga silicone wire to the springs. I should see more improvement once I get that emitter replaced and fire it up again. 
Be very careful without having the springs braided and dont try running it on any less than 4 cells. that driver/FET without spring mods is very likely to heat up / melt springs. I always do the springs first thing on almost all my lights.
Thanks, CK. I haven’t had luck with any braid attempts in the past. They always fray after a few uses and I don’t trust them. Using silver plated copper wire works as well and doesn’t fray or go bad. I’m sure it’ll be just fine. 
Your scaring me Would the copper wire not break with fatigue? My first attempt has shown no fraying. If I did it right, I used 600 Emory paper to flatten and smooth the tops of the braid soldered to top of springs.
I am glad you were able to get it running, I was beginning to think you might “pull your hair out” trying.
You are using different copper braid than I have. Where did you get that? I’ve been trying with copper braid wick but one, it sucks up the solder and gets stiff and two, the copper is so fine that it frays when flexed a few times. Yours looks like thicker copper strands and it doesn’t seem to suck up any solder except for right where you are soldering.
The wire I use has a very flexible silicon insulation and is copper strand, not solid so it will flex way more with zero fraying and little to no breaking
I do like where you ran your braid to on the PCB. I will think about doing that to mine. Doesn't it interfere with the PCB sitting flush to the tube that way?
I also use silicon wire with removed insulation all the time for springs and it works perfect.
There are different copper braids and the ones which work good for desoldering are fine and have flux in them, but for braiding tailcaps they behave exactly like you said they suck up solder get stiff and break.
is sometimes point where copper soldered. I used .125” Solder Wick but only soldered top of spring to make a pad. I try to allow whole spring length of braid to flex so no problem yet over 5 months. If I see fraying to begin I’ll order the silicon wire you mentioned.
Thx
Wash the braided solder wick in alcohol first, then apply flux only where you want solder. The wick is impregnated with flux and if you don't get rid of it it'll pull solder to places you don't want it.
I finished my Terminator mod using this driver. 3700 OTF! Not as high as I'd hoped but super impressive none the less. Thanks everyone for your input!
Hi, can anyone specify the following components of this board R1, R2, C1 & D1?
I lost D1 and I want to buy some spare parts.
Thanks!
Oh, and on all these drivers that keep the Nanjg's voltage divider layout (the 4.7K/19.1K resistors), the 4.7K is the one that has one end connected to ground. It's useful to be able to identify which goes where in case the silkscreen labels are missing/unreadable/incorrect.
Didn’t that issue get corrected in the V1 releases?
But yeah…the R2 is the one on the resistor bridge tied to ground
Good call sir
The silkscreen is fragile on the Oshpark boards, it's not uncommon to find they're partially scraped off in shipping since they just throw all the boards into a padded envelope. It's best to know where things go without having to rely on the potentially missing labels.
(also, if you can do it without needing the labels that's a pretty good sign you understand how the board works, and that's always a good thing) :)