Modding the Solarforce M8,Now Shaved MT-G2! :)

I just got the now obsolete Solarforce M8. Love it’s style, love the modular approach, but don’t love the power output. Can it be modded? The pill is an external component that is not easily yielding it’s secrets. Before I break out the BFH and go in, I thought I’d better ask some questions.

I somehow managed to get an early version, serial number 89. It has an XM-L U2 and the original 2 tailcaps, one electronic with Hi-Lo-Strobe and one on/off tactical style. I got to wondering about the electronic switch, can some gains be made there? So I opened it up and this is what I found…

So now I’m wondering if it’s possible to simply stack a resistor for more output, or put the tactical on/off switch on and go in for the kill on the light engine. Anybody? No one’s ever done this before? I’m thinking an XM-L2 at 4A or better would fare very nicely in this hybrid reflector. Possibly even de-domed.

Orrrrrr, the venerable MT-G2 at similar 4A output level!!! :slight_smile:

Would love to hear some input on this, can hardly wait!

Thanks,
Dale

Well, when you want something bad enough…get the big hammer! :slight_smile: And that’s exactly what it took! Never seen anything like this one before, and they did not intend for it to be modded. Sorry Solarforce. The white disc filling the top of the pill is actually about 3mm thick, this is the primary reason it’s difficult to get out. I put an 1/8” diamond ball in my dremel and bored 2 holes through the plastic down to aluminum, one on either side. Then stuck a narrow blade screwdriver in and pried it right out of there! The raised lip around the emitter that the reflector seats on quite snugly is actually a separate centering ring that fits snugly into the white plastic filler. Everything fit like it was poured into place.

The shelf on the top side of the pill where the pcb sits is actually quite thin, not much metal there to sink heat which is all too common in most lights. The small pcb board is aluminum, with the XM-L U2 nicely centered and connected thermally with a grey paste, something like Arctic Silver 5. It moved freely once I took the spacers and fillers off. See where I bored down beside the aluminum, barely cutting into the anodization in the pill head?

And here’s where things got weird. There’s 2 driver boards, one with the circuitry and one contact board, not uncommon at all. But, the contact board is on the outside of a brass sleeve, the driver is on the inside of the brass sleeve. So the driver is set deeply into the pill. The sleeve is machine press fit into the aluminum pill. It doesn’t come out. I mean, I broke things and it didn’t budge. So I drove a drill bit through the driver, pushing the contact board out from where it was heavily soldered in 3 places to the brass sleeve. This isn’t a simple ring to make ground, but an 11mm deep 2mm thick brass sleeve that’s in there like it grew there.

Here’s what remains of the driver.

So, looks like I’ll be modding my brand new M8 after all, what do ya think? :slight_smile:

Nice removal job. Thanks for taking one for the team. Sounds like its time for a custom pill or lots of copper stuffed in there.

Glad you're not my dentist.

I’m running a special on extractions, this week only…

The brass sleeve will limit the diameter of a copper core stuffed in there, so do I get a whole new pill made out of copper or just ream out that brass and fill it with copper? I have a 7/8” diameter x 1” long solid piece sitting here staring at me, already mounted and ready to chuck up and turn. I could do it like I did the K3, cut the shelf out of the pill so that my copper will fit through it, then the Noctigon or SinkPAD could sit directly on copper. I’m really liking the idea of making it fully from copper though, make the fins larger diameter integral to the pill, maybe even add 3 or 4 fins in length, that’s the nice thing about this light, I could run 5” of pill in between the battery tube and head assembly if I wanted to. Of course, that’d just be wrong, but adding some diameter and a bit more length wouldn’t hurt the dynamics of the light at all and would add a whole lot of heat sinking.

Now the real question is how much would a full copper enlarged pill add to the cost of the light? Yeahhhhh……………

I can redo some things later for better heat control if necessary, but I really wanted to see one of my XM-L2 T6 5000K emitters in the M8. So, I put one on a Nitro PAD, wired it up with Teflon coated 24 ga wires and put some Arctic Silver 5 under it for now. Works like a charm! I put 2 380 binned chips on the inside of a QLite, it’s pulling .06, .96, and 3.59A in the preferred default settings on the driver. According to Match’s charts with the XM-L2 on copper that’s in the 1300 lumen neighborhood and I’ll take it! Beam color is soooo much nicer, just a white light with no blue or green or brown or anything, white. Just like I like it! :slight_smile:

Not bad for an evenings work! :wink:

Got a surprise, the XM-L2 at 3.59A is actually the same or brighter than the XM-L at 4.28A!!! Usual camera and settings, Canon G1X zoomed in to 112mm at f/5.8 .5 sec 1600 ISO. M3 with XM-L U2, de-domed, at 4.28A first

Now the M8 with XM-L2 T6 at 3.59A

And here’s the M8 yesterday, with the XM-L U2 stock at 1.66A on 2 Samsung 20R cells. Which is another point, it’s now running only on one…no more 2 up config.

Opening them both up, then switching back and forth on my 24” Dell Professional monitor it’s difficult to tell em apart, looking at them the same way…full size…the M8 is a bit brighter. The M8 shows a bit more blue in the light than the de-domed XM-L in the M3. But they’re very very close! The M3 with the de-domed XM-L heats up faster at the 4.28A it’s pulling.

Win! :slight_smile:

Is a higher vF is pulling similar watts with the lower draw?

The XM-L2 is supposed to be more efficient, so 3.59A to 4.28A might be that much difference. Might also be some visible difference due to a more efficient reflector in the larger light.

Doing some quick math, a 20% gain in the emitter would be almost exactly the difference in those 2 Amp figures. So it could be as simple as that much better emitter in the L2.

So I added 4 more chips to the QLite in my M8. Running a Samsung 20R cell it’s hitting 5.02A on a fresh charge! :slight_smile: Will try to get a beamshot, weather permitting.

I’ll repost the stock, first mod, and now this mod to show the gains. 1st is stock, second is at 3.59A with XM-L2 on Noctigon, third is at 5.02A retaining the XM-L2 and Noctigon.



These should pretty much speak for themselves. 97 yds to the red oil drum from the camera. All same settings taken with G1X 1600ISO f/5.8 at .5 sec exposure.

I’ve gotta bring this one back to life just to say thanks! I’ve considered driver modding both the m8 and m6 for nearly as long as I’ve owned them but didn’t want to run the risk of destroying my two most expensive lights. Now that they have some age on them and the xml2s are available I’m starting to reconsider. Do you think it would be possible to cut off the shelf, press the sleeve out, and press in a thicker copper or aluminum sleeve that’s been threaded to accept a p60 pill? This way a person could have easily swapable led/driver combinations that fit both lights (as well as conventional p60 hosts). If the sleeve were threaded all the way, a person could make up some taller copper pills with p60 compatible threads and more mass/contact area for the high current stuff without sacrificing the ability to use conventional ones.
I’ve got a piece of aluminum tubing sitting in front of me that looks close to the right size. I just need to figure out what size tap to buy and work up the courage to rip into the light.

Tony

I just got through putting a shaved-dome MT-G2 in my M8, on laptop pulls at 3.74A I’m getting a start-up 2138 lumens. :slight_smile: had to tone it down, was running the MT-G2 in Direct Drive at 7.51A on those pulls, 12.18A on the 20R’s and it was way too hot!

Sounds like a good plan to me, give it a go, I’m betting you’ll be glad you did!

I’m interested to see how you shaved the dome and the resulting beam profile and tint change.

I used a brand new Xacto knife blade, with the pill of the M8 for a guide…shaved it to about 1/32” above the phosphor. Used 1000 grit paper to smooth the surface, it’s not shiny but it’s at least smooth enough to see the dies. The hot spot is tighter, still white, with a warm spill. Pretty nice actually. Throws better too. Mind you this M8 has a hybrid reflector, with orange peel the first 12mm or so and smooth reflector the rest of the way, pretty deep reflector.

And yes, it cost some lumens out the front. I have more 7135’s coming, so I’ll add 2 more pretty soon and get the amperage up from 3.74 to around 4.2. We’ll see how that works out. I really don’t want to make this one a blaster, as the pill is really too small for the heat that’s made at higher output levels. As soon as I can get some funds cleared I’m going to have a copper pill made for it, a bit larger diameter and about 12mm longer. Then I’ll bring it back up to 6-7A.

Edit: I did take Comfychair’s advice about using silicon to lube the blade, worked out pretty well. :wink:

Wow! I can’t wait to get a chance to play with an MT-G2. I think that might be a perfect match for the already floody Solarforce MPP2 head if I can improve the thermal path some.
The M6 and M8 are my go-to lights for reliability so I’ll probably go with conservatively driven XML2s in both to start but put together several p60 pills for them as time and money permit. Thanks again for being the first (that I know of) to open up the M8 and post about it!

Tony

Hey Tony, be prepared to destroy the OEM driver. That white disk that is over the LED is actually fairly thick, it has thick outer edges where it meets the pill and the only way I could see to get it out was to dremel a small hole through the plastic at the edge, (you might go about 2-3mm away from the edge) and then use a small jewelers screwdriver to pop it out. It’s a tight press fit but does come out…I put a hole on each side and worked it up, you can see where the wires are on the star so just put your hole opposite those.

After removal of the star I used a small drill bit to punch through the driver and push the contact board out. They made that press fit brass sleeve with insets on each end, the driver is inside and of course will not push through the sleeve. The contact board is soldered to the sleeve first, then the driver is press fit on the inside making a sandwich. Then the entire assembly is apparently pressed in.

If you bored through the shelf where the emitter sits, you probably could press that sleeve back out. I just used it to solder my copper heat sinks to… made 2 discs from 1/4” thick copper, large enough diameter to fit snugly down into the pill and then re-flowed the entire pill to get the copper soldered to the brass sleeve.

It’s not like any other light I’ve seen, but I really do like it’s form factor and now have it running pretty much like I envisioned when I first saw it. :slight_smile:

Just gotta stick with it! :wink:

Thanks Dale! Before I get started I’d like to ask a machinist friend about your idea of making a custom pill. Something with more mass, an increased number of taller, thinner cooling fins for more surface area, and the right id and threads to accept the p60 pills. I might also see if I can get him to thread a piece of copper round stock to match. Then I can cut pieces off of it and drill wire holes and driver pockets myself to make my own p60 pills.
I found out last night that the lip on the store bought pills touches the bottom of the M8 reflector before the led is deep enough. It wouldnt be a big deal to cut them down but I’m hoping I can make some with flat tops that will work with the M8 and still have enough thread engagment in a p60 reflector.
As usual for the planning stage of a mod I’ve got a bunch of ideas about how to proceed and just when I think I’ve settled on one, a new one pops into my head. I’ll pick one and get started eventually. :slight_smile:

Tony

Thought I'd go ahead and get a couple of beamshots and compare it to my new Convoy C8 while I was at it. The Convoy has an XP-G2 R5 2B and is pushing 666 lumens out the front using a Qlite at 2.95A. The shaved-dome MT-G2 in the M8 is pushing 1638 lumens at 3.74A on a Qlite with 2 chips added.

The set-up, my standard fare...The G1X in manual mode, f/5.8 at .5 sec exposure, ISO1600 with 2 second timer and manual focus. The red oil drum in the center of the picture is 97 yds away. Wide angle shot is 28mm, zoomed in shot is 112mm.

The M8, wide is base, mouse over for C8

The M8 zoomed is base, mouse over for C8 zoomed

I will be bumping output on the M8 when more chips come in next week. :)