Torchlite SVD7, De-domed MT-G2, 9A driver

Flashlight Modding and DIY Parts
1

I started with this host
http://edcplus.com/torchlite-svd7-2x-26650-diy-flashlight-cree-xml-sst90-sst50-p-127.html

this driver
http://www.intl-outdoor.com/9a-3mode-5512v-circuit-board-p-561.html

and this emitter
http://cutter.com.au/proddetail.php?prod=cut1381

Emitter:
*Sliced dome with razor, polished to 1000 grit and then one drop of clear coat added.
*Reflowed to sinkpad MCPCB.
*Dremeled notches to accept larger wires and screws.
*Sanded bottom side of MCPCB flat to 1000 grit

Pill and driver:
*Drilled/tapped pill for 4x40 screws
*opened holes for leads up to take 18ga wire
*Sanded <2mm from interior of pill side for driver clearance (400 grit)
*changed leads to turnigy silicone 18ga
*Potted top board of driver (devcon 2 ton epoxy and alumina)

*potted bottom board of driver

*applied release agent to inside of pill and potted entire driver inside it

*soldered driver to pill

*applied arctic ceramique to pill and sinkpad, assembled pill with nylon screws (probably change to steel when I find a suitable nylon washer)

*trimmed insulator

2 cgr18650ch feed 6 amps at the tail (*edit, this may be inaccurate as cells were not fully charged), and that drops to 5 amps within 30 seconds. Instant-on lux >22k@1m.

Beamshots and more current/lux measurements will be posted when my MNKE 26650s arrive.

2

Good lawd! That’s a hotrod. How did reflowing the emitter onto an XM-L Sinkpad go—was it more difficult due to the different pad size of the MT-G2?

3

Wait, 5+ amps at the tail? You realize that MT-G2 is getting seriously overdriven (like an insane amount) at 7 amps or so... Great work on the mod though, very neat.

4

Wow, if Torchlite could have had these ideas for their light, they never would have gone out of business. Your project is loaded with top of the line everything. I love the potted driver too. The light OTF must be truly something to see.

Amazing job!

5

The sinkpad is designed for an MT-G2. The footprint of an XM-L emitter is way smaller.

6

I was wondering how the heck he’d managed that with an XM-L PCB and not shorted +/- on the thermal pad…

I was under the impression we had to do group buys to get stuff from SinkPAD. This is great news, as I want to get my hands on their MK-R boards…

7

reflowing was much easier, because it is big. I have a heck of a time reflowing xp-e/xp-g.

I put the pcb in a clamp and touch a 250 watt soldering gun to the bottom. I just blow on it to cool it off. The (thicker) sinkpad took noticeably more time to heat up than the cutter pcb.

As you can see by the brown stains on the board, I use flux. The brown stains on two corners of the emitter are where I pressed down with my tweezers to squeeze out any extra solder/flux.

Yes, I’m aware that I’m roughly doubling the cree specs for current, but somebody had to try it :stuck_out_tongue:

When I open it back up to put steel screws in, I’ll check current to the emitter, mostly because I want to know efficiency.

The beam is interesting - a fairly well defined, even, large hotspot, and the spill might be more accurately described as one big corona lol

Thanks for the compliments - quite a bit of time went into that…

8

yes, that is an mtg2 aluminum sinkpad.

I’ve experimented w/ xre on xml sinkpad, and you can be sure the mtg2 would die quick death on an xml pcb with this kind of current.

the sinkpad is nearly the perfect thickness for this light - the reflector rests on the pill and there is only a little adjustment available for focus (as the oring compresses).

9

I wish I had realized I could have gotten a little extra height on my mtg2 with the Sinkpad star being so much beefer or taller. I wouldn’t have had to shave the top of my pill for additional focus. I’m going to have to give reflowing a shot and make use of these stars that are sitting here.

10

I hear you, I’m near sweating bullets when reflowing Nichias even though I’ve reflowed half a dozen XM-Ls.

Sounds like my kind of spill; I’m not the fan of pencil beams that I was when I started this hobby. Looking forward to beamshots 8)

11

Very nicely done; you should be proud of that light !

12

thanks! Hopefully someone agrees enough after the beamshots to buy it, because I have two more to build :slight_smile:

I did some late night testing with fully charged batteries, and I might have been off with that current reading. I think the cells were at a lower voltage than I thought. This is why I need a good power supply!

I won’t confuse the topic further by stating more numbers until the 26650s arrive.

13

Nice job. What did we say when we started discussing these hosts? That should properly trounce most 3 x xm-l lights. 8)

I took my civilian sedate version of this out last night. It is a very usable light it has to be said. :slight_smile:

Anyone thinking about getting one, just do it, they are a really nice build and the host quality is second to none.

14

Awesome! That is an amazing build you did there dthrckt. Thank you for sharing. Can’t wait to see the beam shots and hear what it’s like to use in the real world.

15

last beamshot in this series, is the light in the OP

16

It was used earlier but I'm using it to. WOW. What a build. To Be honest I'm gob smacked. Cant wait for your 26650's to turn up and night shots taken.

17

Thanks!

update

I found 4x40 nylon washers and switched to steel screws. This made a big difference in heat sag

The mnke 26650s showed up yesterday, but were not done charging when I took the beamshot, so I used charged cgr18650ch. Those fed 3.8A at tailcap.

When the mnke 26650s were finally charged, I did current and lux measurements again.

4.2A at the tail. 24k lux (converted from 13ft to 1m) at turn on, which settles to 21k after a few seconds and holds.

I need to do a little work on the screws. The head is thicker, plus the washer and I want to be sure they’re not interfering with focus adjustment.

Then I’ll do whitewall beamshots, and a better night beamshot, but that probably wont’ be until next week because I’ll be gone this weekend.

18

That last shot shows a warmer version of what I see with mine.

Its not too impressive, till it occurs to you that at 75 - 100 yards the HOTSPOT is lighting up at least two terraced houses lol 8)

19

yeah, I think de-doming made it noticeably warmer, but only added a bit to throw.

I think I’ll have to bring out the DRY for the next beamshots because comparing this to other 2 cell single emitter lights is comical inside of 100 yards. there’s just light everywhere lol

also, it looks like the driver is putting out the stated power but is not truly constant current (sst-90 at 9A would be about 30W)

so, when I measure current to emitter, I expect to see 4A (~75% efficiency) to 5A (100%).

Fully assembled current might be a hair higher due to resistance of my meter leads?