I’m not even going to attempt desoldering everything…what a nightmare. If it works…fine. If I find it’s not delivering, which I doubt, I’ll build a new one from scratch.
Nice come back OF. If this was my light I dont think it would still be a work in progress. 
I’m the comeback kid. I have to be, as I’m always screwing my mods up.
I tested my driver and it is now installed in the light and wired to the MTG. Just have to complete the carrier and wait for my 18500s to arrive...cross your fingers it all works.
I'm using an HX1175B buck driver.
With two 18650 it drove an MT-G2 at 3.5A, and with three cells it drove the same emitter at 7A.
Why the difference in current readings? I've used LCK 5A buck drivers with MTG and two cells and got the full 5A. What is different about the HX1175B that it won't produce 7A with two cells.
I'm not worried about it, I just want to know how it works.
I cross posted your question over here. In the meantime, can you measure LED vF while driving it on two cells? Ideally we’d also know the total loaded voltage of the two cells. I’m thinking that maybe this emitter has an exceptionally high vF or the cells are sagging exceptionally low. Vf + sense voltage (0.25v) must be > input voltage. If Vf + 0.25 = loaded battery voltage then there’s your culprit. Really it’s just something to check, I don’t know.
I still remember when you first purchased this gutted host and wondered if you’d ever attempt such a mod. All I can say is, “WOW!” :bigsmile: Great job Ouchy!
The 18500 carrier is finished.
I used Fujik to glue down the switch. It sits right on top of the neg. pad, almost covering it. Just a hair is exposed. Drilled a hole from the neg. pad of the first cell in the series, and soldered a wire to the switch. On the other side, I soldered a wire connecting ground to all three brass posts.
I felt like fiddling about with the carrier, and wanted to improver the resistance where the steel nail heads connected ground. I wrapped copper wire around them, soaked all in liquid flux and let the solder flow. The solder bonded the wire to the copper plating under the nail heads, and flowed through to the brass rod. Then I ground down the heads some more, exposing the copper. This should help to improve ground resistance.
Of course, I couldn't wait to see if the darn thing works in the flashlight, so I loaded the carrier with two Efest IMR18500 (all I had) and a 14500 dummy.
It works!
With the light on, I placed my probes on either side of the switch and got 3.5A, which is what my initial test of the driver with 2 x 18650 gave me. Thee cells should give me 7A.
I took some crappy beam shots with my Ipad running just two cells. It's high noon, and my apartment isn't very dark, so I can't really say much about the beam.
Here is a wall shot from about a foot away. Note the dark spot in the center. It disappears when the light is pulled back. What's the story behind the dark spot and focus? I forget.
The tint has no green in real life. My ipad camera thought it would be nice like that. I saved it to show the "dark spot".
The next two pictures were taken instantaneously. One snap...two pictures. One is regular, and the other HD. (It's still just an ipad camera though)
I think if both photos met half way, it would be close. The hotspot of the second pic with the spill of the first.
Like I said, I'm only running at half current, and it's daylight. I haven't got a handle on what this beam is doing.
It seems to me, that instead of the usual MTG flood, I'm getting a hotspot with spill.
I'm just waiting for my three Sanyo 18500 to arrive so I can power it up full.
The heat sink is doing its job. It takes a while to transfer through all that copper. I think it just keeps absorbing the heat, then the battery tube starts getting hot. I knew this was going to happen. I had it on high for a while trying to take some photos. When I removed the battery carrier, the cells felt toasty. Not a light for a non modder to play with.
If you have a lux meter, you could still measure lux, even if it’s not night :)…
Sorry ohaya, no lux meter.
I really want to get it outside for a visual. Since the original reflector wasn’t designed for LEDs, the beam profile will determine whether or not the entire build is a fail. So far it’s looking like it has a nice hotspot. I’d like to compare it to some other MTGs.
It looks good to me man, I don’t think there’s any way this could be considered a failure.
Just an FYI I found a SANYO 18500 brand new in box (and at ~4v) in the recycling bin once and since I don’t have any 18500 lights I use it as a test bench cell, it easily does >5A direct drive and doesn’t suffer much sag. If your looking for other cell options definitely consider them.
I just tested 2 x Efest IMR18500 direct drive with an MT-G2 for ImA4Wheelr. They drove it at 6.9A
I’m expecting three Sanyo 18500 by the end of the week.
Great effort, but u guess in the end when it workit makes you feel very happy. Can’t wait for some outdoor beamshots.
When I clicked the switch and light came out the front end, a lot of weight was lifted from my shoulders.
I could here the whew from here. Nice comeback and well done. 
Yea its a very good sensation especially in a difficult and long project, when the time come you click and there is light 
My three Sanyo 18500 arrived. The MT-G2 is now running at 7.5A. I was testing it into the little forest beside my balcony, and it suddenly stopped working.
I removed the battery carrier from the light, and all the batteries fell out. I guess the springs didn't like that much current and they collapsed. They weren't the heaviest of springs, so I was kinda expecting that.
BEFORE
AFTER
Other than that, the light works great.
What a accomplishment Ouchyfoot. Well done to the max :-)
Threads like this just makes me happy :-) TY