Thanks for the detailed infos on your mod. I will just do the same.
I never used a shim until now, have you bought this piece as a shim or just cut it out from a bigger part?
I have a copper sheet but it’s not thick enough, maybe I have to do a little search in the scrap parts box…
Edit: just searched a bit and there are tons of shims on ebay.
Werner - post above updated with link for the source of copper shims I used.
Hey Tom, that’s what I’m doing next is the blf17dd driver on top of the switch like your ld-1 driver mod. 
Been busy (putting up more solar)- but my drivers came in the other day.
I’ll piggy back it, but I’m not too sure where to put the eswitch, I was thinking run it as a temp on/off button or is there a specific place on the board to let that happen? I really need to get learned on flashing these things too but one step at a time.
The stock driver noticed on the Y3 is a lot more tanky than you’d think. I know wight was saying the diode is rated at 5a but I’ve ran that U2 1D as low as 0.036 with 2 cells successfully. It heats up like crazy though, not really practical. If you swap the R150 with a R120 and lay the R100 on top, keeping the R180 in place, it should give you 6a and still runs cool. I can’t measure it cause my meter doesn’t work at the tail after 0.45. 
You were right about the dedomed 1A. It is a bit more white than the 1D after dedome. The 1D is no slouch though, but I like the 1A better after the domes off. Playing with the depth in the reflector, I wasn’t being precise like you either. I should measure the centering ring after it’s sanded down. I’m just sanding and putting it in until it’s centered and I see that flower look. I noticed if it goes too far into the reflector you actually lose throw light. The beam gets tighter but not as bright? So I stick it back a little but not quite the thickness of the stock ring. It’s probably exactly where you have it sanded, wouldn’t doubt it anyway.
That mcpcb is perfect size… I can already see the cousins asking me to make them a few of these after my inlaw blinds a few of them. I’m not looking to my cousins kid though he always plays with the deer call in the stands, thinks it’s funny… And then there’s the hillbilly drama
Interesting fellfromtree... I was thinking of boosting the amps a little more - I'll do the resistor math and see. for the focusing, actually I didn't do any experimenting - should have. The throw is good, but hoping for a little better, but I'm not measuring ampds to the LED with 2 cells, so hard to tell. Are you having problems fitting 2 protected cells with these extensions?
I had to take off the button tops to fit protected. You can add o rings found too but that’s last resort sort of thing.
If you play with the depth on the dedome on a white surface about 20 feet away. You can see the light definitely loses throw at a certain point. Like it loses intensity- this is just with the naked eye though. The beam gets more floody if you go all the way in, but that’s what I noticed anyways.
Edit
If take a picture it looks different but you can see almost to the point of where it starts to get square… It looks perfectly round in rl. Maybe I need to play around with the camera some- but it has these 4 petals that come off the hotspot… Right at that point you have the richard simmons sizzler!
thanks tom e for the details, smaller quantity shims : 5 X Kupfer Kühlkörper Laptop GPU CPU Kühlkörper Kupfer Shim Sale - Banggood Deutschland sold out-arrival notice-arrival notice
Im watching this thread.
even a little cheaper at ebay:
sets of 10 pieces:
20x20x1mm = 3.62USD
15x15x1.2mm = 2.80USD
etc.
The eBay source though for qty 45 in 9 different sizes is still the best deal, if you want the variety of thickness's.
Instead of shims you can grind the base of the heAd down instead
I looked at doing that - won't the bare aluminum be exposed and seen? Specially if you have the light sitting head down? If so, does it look ok?
tome did you solder the shim with the led pcb ?
Yes - it's noted just above a pic.
I just took some measurements of LED current. The Stock current is between 2.4-2.6A over a wide voltage range. Power input is around 10W.
I will add a resistor and take measurements when I get the retaining ring out. It’s stuck and I used some oil and will let it cool down before I try again. The holes are a bit too small to use my good pliers and with tweezers it was unscrewable…
Edit:
I just played around with the driver. I soldered the driver to my test xml2 and played a bit. On stock there are one R180 and one R150. One added R100 boosts the led current to around 4.3A. Input power is around 23W
I also tried for a short period another R100 which gave around 6A to the LED at 35W input. After that I shorted something and needed to replace some parts so I couldn’t test the 6A for a longer period. After the repai the replaced FET wasn’t capable of the high current and I removed one r100….
I am now sure that the partly defective stock driver isn’t powerful enough for me in this light so I will probably also use a FET driver piggybaggedn one cell config.
Now I need to order a imr26650…
Thanks for the test results Werner. Did you record any output voltages?
Based on djozz’s Vf data I’m assuming >80% efficiency in stock form, down to <70% with one R100 added and <65% with two R100 added.
I didn’t took measurements of the led voltage(yeah I know that was lazy…) and you have to consider that I used a different led(noctigon on big heatsink) for the r100 test. The led voltage on the stock measurement was probably a bit higher than usual because the led was in the stock not ideal pressed down and got so a bit hotter. Also the input voltages are not super accurate so the data looks a bit worse than in reality I guess(that is why I didn’t post a lot datapoints). The regulation is not constant over the whole voltage range there are 10% fluctation like always on these buck drivers…but it is good enough to get an impression of the driver.
I also like a lot that this driver has voltage protection for one, two and three cells.
The efficiency goes down for sure a bit, I would also say efficiency is around 65-70%.
Think'n for my R100 moded light, add an ~R200 - think it would be a nice compromise. Also I have to play with the focus on that light a bit. Just wish I had time...
Another r200 will give around 5.3A.
4A in this light would really be a bit low for these days, especially because the heat sinking capabilities of this light seem superb.
Does anyone know a cheap source for the keeppower imr4200?
Looks alright to me! Did give it a quick rub with a finer grade and clean up the edge a bit.
Hhmm - yes, doesn't look bad, litle hard to tell form the pic. I'll have to give that a go.
Werner - sorry, no clue bout where to get the KP IMR 4200. I'm using the KK 26700's - got 8 of them.
My first Y3 I modded, used a DD driver (not sure if Nanjg w/FET or BLF17DD), and it was awesome - 1,700 OTF and bout 135 kcd, with a UCLp and XM-L2 U2 1A, and think I was only getting bout 5.6A.
I may not go that route, but it definitely looks like a good and attractive solution! Thanks for posting the pic.