I had a nice project in mind today, hotrodding a ToolAA 2.0 with a 15mm BLF-A6 driver and LH351D. But instead I spent an hour getting the pill screwed out, with brute force, and heatinh the head up with a blowtorch. While applying more and more force and heat, I first broke out the stock driver, then damaged the led shelf, then after a particularly intense heating session, the reflector coating started peeling. The pill never moved.
So that was one ToolAA down, a very rare occasion that I was not able to get a glued part loose. (this used to be a smooth reflector)
Using Aluminized Copper Reflectors as Secondary Heatsinks
Was wondering why aluminized copper reflectors are always thermally isolated from the emitter star, when their composition and construction might so conveniently lend them to use as secondary heatsinks. Yeah the base gasket does help keep the reflector centered correctly around the emitter and isolates it electrically, but the lamp barrel holds it in place anyway and on a large star with + and - solder points nearer to the starās edges risk of contact with the reflector is basically nonexistent. Thinking if the gasket is replaced with a well trimmed thermal pad and the reflector rim is similarly treated with paste, then heat from the emitter could travel unimpeded through the reflector all the way to end of the lamp barrel and into the bezel.
Anyway as proof-of concept unscrewed the bezel on a Sofrin SP33 V3.0 and removed the reflector and base gasket from the lamp body. Next (because my .5mm thermal pads have not yet arrived) applied some no-name thermal grease to both the base and rim of the reflector, lowered the lamp body back down around it, then flipped everything back up. Before rescrewing the bezel with its O-rings and cover glass, fired up the Sofirn on Turbo to see if I could feel any heat in the reflector this way. And boy did it get hot! So screwed everything back in place and let the flashlight cool down against the window while readying my second SP33 for comparison. Turned both units on in Turbo simultaneously while clasping them in two hands. Thermally modified Sofirn got so hot around the bezel I couldnāt hold it long. Unmodified unit got hot too but not as painfully so. Thus I declare the experiment a success.
Now, none of the reflectorized flashlights or bike lights in my possession have any problem whatsoever with thermal regulation/step down, so this mod is not all that useful to me. Then again my Kaidomain BL70S XHP70.2 4000K bike light doesnāt get hot at all on its highest setting of 3000lm ā but maybe it should? So once I get the .5mm thermal pads will see if they make any difference with thermal transfer in that lamp. JKK36R has similar build so might be useful there too (though Iāve never used turbo on that flashlight for more than a couple of minutes, out in the cold.)
Yes, this is a significant untapped source of thermal capacity. To my knowledge Armytek does it already but I havenāt seen that elsewhereā¦until todayā¦thank you for sharing.
i have this zoomie also, i changed the led several times , from fake xhp50.2 fake xhp70 and a funny looking star with 4 leds each with there own dome on them
nothing was a powerfull as the original fake āxhp90ā that came with it
now it only goes into low mode and is wierd, i think one of the smd resisters walked away somewhere
cost 30 euro including postage to finland
didnt come with a charger as promised
refunded in full
well ive got a zoom host and 2x 26650 batterys out of the deal for free
i might put a lexel driver in there and a xhp50.2 and see what it does
I have a few lights where coupling the reflector, thermally would be possible. The XintTD C8 actually has it designed in, if it will transfer that way. Adding the compound should certainly make it better.
Now, the main concern I have is how messy will it get with some arctic silver goo(?).
It will come back at some point
To get over it I repaired my Maratac AAA Copper. It had a 4000K 80CRI XP-L2 on a 10mm Sinkpad and a Mtn direct driver with guppy UI which was wonderful, but I never did the ledboard soldering right so eventually a joint came loose. Now I used a better DTP board (from vestureofblood) and a 3000K 90CRI LH351D (got it from member contactcr), and renewed the ledwires, did a much cleaner job now. Iām not sure if I like it better than the XP-L2, but it works again and will keep working this time
I have experienced it a few months ago when trying to swap the led, ended up with damage reflector, LED gasket, and glass.
I then contacted Lumintop Aliexpress to ask if I could buy the parts only, and they said yes! Just say what you need and they will sell it for you. Hope this helps to rebuild your Tool AA again.
If you canāt get a reflector from LT, send me a PM.
Also, if you want to use a nice TIR, here is the pebbled TIR I am using in the Tool AA from Henk4U2 with the Luxeon V2 3000K.
I put four optisolis on an XHP70 mcpcb. The terminals needed to be widened. Its the hardest reflow Iāve done yet so Iām proud of it, even its solder blobs.
What gauge wire should i buy to bridge these and the VR16sp4?
Is that blue stuff also under the board, to electrically insulate the ledboard from the pill (which is ground)? Iām not sure how hard you will drive the leds, but over say an amp you may run into thermal transfer problems through that thick layer.
Yesterday I finished modding my old Jacob A60 clone with a CSLNM1.TG (white flat 1mm) driven by 12x7135 nanjg and a forward clicky. Possibly going to air wire the driver so that I can have a star configurable single mode. Iām seeing a consistent 235kcd with this setup. Reflector tuning on this light is notoriously finicky, and I get to choose between beam artifacts and losing ~15kcd when swapping between different centering rings (clear vs white). Right now I chose having the artifacts. Pretty crazy the throw to size/weight ratio here.
Before installing the driver, I was running it direct drive off an old LG E1 with elevated internal resistance. Charged to 4.35V I would see 5.4A max (where I saw 250kcd turn-on) and it would stabilize at 4.4-4.8A for a long time actually. I feel now that Iām running the 7135s, there is a noticeable increase in heat generation.
EDIT: The Osram replaced an old XPE2 R4? dedomed running 2.3A direct drive from on old Sanyo UR laptop pull. That setup did 150kcd turn-on, ~140 after settling in.
Fell 3 inches and removed everything Itās a Manker E02. I was successful and reflowed a new LED but it was misaligned and reheated it. I should have just taken the victory.
Anyone have any ideas how I can start fixing this? Most of the parts need to be reattached to the back(between the MCPCB and copper heatsink).