Mod: Small Sun ZY-T08 Emitter Upgrade - MT-G2. Now you can call it a Small Sun

Have you done a comparison between two otherwise identical lights, one with a plate of sufficient thickness blocking off the hollow pill but no other extra mass added, and one with the full coppercoppercopper treatment?

Yes, that's a leading question, as I think I already know what the outcome would be... though, if you HAVE done that comparison and found the 'mass for mass's sake' method showed an improvement, I'd (honestly, truly, not sarcastically) love to hear the details.

If your kitchen sink routinely overflows and spills all over the floor, the best long-term solution would be to:
A) Buy a bigger sink, to extend the amount of time water can run before it starts pissing on the floor
B) Improve the drain, by checking for and then fixing any sections that create too much restriction to flow

I actually think sometimes, a heatsink in a light can be counterproductive. You don't want to absorb and accumulate heat anywhere inside the light. Get rid of it. Mass above and beyond the amount required to carry the given heat load not only doesn't transfer heat faster, it can actually slow it down. In the hot-rod PC overclocking world there is such a thing as a heatsink with a base that is too thick, and a thinner base with no other changes to the design can give better performance. The added mass will smooth out the rate of temperature change, but that also increases the amount of time that heat is retained after the load is reduced.

You only need enough mass to serve as a wide enough path to move the heat load (2-lane backroad vs. 8-lane highway (sorry, this is the day of corny metaphors I guess, deal with it :p )), anything beyond that, at best, isn't needed.

(I know you're no dummy with this stuff, but sometimes even smart people do things the way they've always done it without really thinking about why often enough)

;)

I think I understand you, and we basically agree the higher end CPU heat sinks have it right - copper close in, aluminum mass with fins on the outer layer. Copper does the job of pulling it, aluminum moves it quick to the air. Now it's true what you are pointing out - if you retain the heat close in, that's a bad thing. That's why I want to be sure the copper is bonded well to the aluminum housing that surrounds it. Typcial aluminum pills are not ideal - I'd prefer a copper pill tightly fit/bonded to an aluminum outer mass, though percentages, portions, masses do matter and will make a difference. I don't believe a SinkPAD bonded to an aluminum pill is enough copper, but I have no data to back that up.

No, I'm not doing the stuff Michael at OSTS does with all the thermal measurements and tweaks. Im' envisioning the copper heat sink on the other side of the aluminum does act as a buffer, but the aluminum of the pill top and where the heat sink is bonded/touching the outer body of the pill will push heat out. Is there some back-wash going on? Sure, and probably that effect should be studied, tested, researched. I'd be very curious myself in finding out and would adapt accordingly. I'm not adding hugh amounts of copper either - it just seems it's commonly done by BLF'ers here and I'd be hesitant of changing until I get some proof/evidence otherwise. I've learned a ton of stuff off of BLF and have read through these debates before and haven't found anything compelling enough to prove otherwise.

Now on the other side of the issue is perception - from the user. The "ouch" effect of this custom light getting too hot to handle - yes, we know that's a good thing, but the impression is usually all negative.

I think at a certain drive point, a light will not be able to get rid of heat without active cooling. At that point, after enough time it will be too hot to hold.

That is where mass comes in. Extending usable run time. Its not like these lights have a 24/7 up time. My computers hs/f has to keep the heat down for extended lengths of time, but a flashlight on high? As long as the batteries die first I don’t care :stuck_out_tongue:

coughNITECORE!cough 0:)

Relic. What is the diameter of the stock driver board?

Driver diameter: 30.1mm.
More dimensions in my review here:

Not sure how I missed this, but well done sir.

Now you have me all set to try this build. The conversion of parallel to series would be a new test for me. I’m pretty sure I follow you, but getting the actual light and mentally imagining the current flow with little arrows moving around in my mind is the best way for me to grasp these things. What I want to know is, will this driver fit into the pill.

It's shorter than the IOS 3-18v (15mm vs. 17mm), so it should.

That driver should fit up in there (assuming that 216mm diameter given is a typo :smiley: probably 21.6mm).
I have a 26mm x 18mm driver in mine, with the original driver still intact as a contact board.

Sorry, I should have checked your original revue. All the specks are there. Yes, I think I will try this. I hope you won’t mind if I get stuck and need to send you some pics to tell me what went wrong. If you were to do it over, is there anything you might do different? Do you think one of those small brass buttons used as positive contacts instead of springs on contact boards would make a good alternative to a solder blob.

A brass button might work, it just needs a slope on the ends so the rotating cells can slide over it. I found that a solder blob gave me the contact and shape I needed all in one. I sanded the top of the blob flat so it makes a little better contact with the cell negative end.
I’d prefer to not have to put the cells in series, but I didn’t have the choice with this driver. I also would like to find some kind of centering spacer for the emitter,a s I cannot get things perfectly aligned as it is now.
Feel free to ask questions and post pics, I’ll try to help out. :beer:

Make a centering ring...

Find some scrap thing (dead TV, coffee maker, whatever) with a flat slab of plastic about .040" thick you can cut a section from. Mark a center point and scribe out a line that matches the hole in the reflector. Drill a ~3/8" hole then make the corners with a square file. Sand/file down the area outside the scribed line by ~.015" (MTG2 square base is .020" thick). If it turns out right that should give you a standoff of .005" between the reflector and the top of the substrate when it's all stacked together.

That’s a great idea… my hand fabrication skills are very limited. Worth a try though.

If the hole in the reflector is big enough you can make the spacer from two .020" pieces glued together. But the hole has to be big enough for the top layer to be a complete circle around the OD - like with the typical XML-size reflector hole, the raised section is really four little... ellipses? Arcs? Whatever that shape is.

I’ll have to figure out the emitter centering when I get there. If you come up with something simple…let me know. Seeing as the MT-G2 needs six volts, you really didn’t have a choice but to go in series. There really isn’t any choice out there with two cell side by side series lights. It’s either convert this one, or you have to go with the long, two cell tube. Maybe I’ll use a dremel with a felt pad and knife stropping compounds to polish up the tailcap. Gold(course) - green(fine) and red(xtra fine…jewelers rouge).
Okay…I’m off to FastTech…if I can pull this off I’ll feel more like a modder, rather than just an assembler.

Not true. The TK35 clones are side by side with series connected cells. Look them up. JohnnyMac did a review on a UniqueFire branded one last year. I believe their reflectors are only C8 sized though.

EDIT- Here's a link: http://m.tmart.com/UniqueFire-UF-F10-CREE-XM-L-T6-10W-1000LM-6-Modes-Flashlight-Torch-Black_p158534.html

UniqueFire F10 is the model.

-Garry

Ouchy, I picked that particular driver because it claimed operation down to single-cell voltages. Once I got it, it only operates in low mode below 4.5V. That’s why I switched the cells around. It was supposed to be a much simpler mod; didn’t work out.

On a side-note, the fate of this light is undecided. I might harvest the emitter for a Mag mod and put a dedomed XM-L2 in here for a pocket-thrower.

Well, I’m committed now. I even have a germ of an idea forming as to the “blob” alternative.

http://www.cnqualitygoods.com/goods.php?id=1178 + suitably modified 50-52mm reflector… or use 3 MTG2s wired in parallel with the IOS 9A driver. :evil:

The LED really deserves a big reflector, though.