LED's are often rated at 85C, so for the electronics, it's probably safe, but still as Jason says, even if I set the limit at 90C, the algorithm kicks in sooner, and just because the MCU sees that's temp doesn't mean anything else is even close.
Again, 90C is the max allowable but you can set it to anything you want. At 70C, I got lights that drop output and are still comfortable to hold. The battery is basically still room temp when it trips.
Nice, real nice. I may consider a 3500K swap as well as its actually my favorite tint. Was it easy enough to remove everything and get to the LED board? Any noticeable boost from the spring bypass?
Sorry, didn't take a before/after for the bypass. The 3500K's look great to me as well.
for the LED swaps, I wouldn't call it easy unless you knows how to solder pretty well. The main LED wires are short stubs, so the driver can't be pulled out without first desoldering the LED wires. The most challenging part is re-soldering those LED wires because in mine, they were barely long enough and the contact pads are very close to the hole, so the risk is grounding out the wires or solder to the main copper of the MCPCB. The 2 LED wires and 3 AUX board wires are a tight squeeze thru the hole of the MCPCB. I made sure I took pics of the AUX board before taking it off, just to be sure I got the wiring right. Only thing I'm noticing now is a don't have a well formed hot spot - it seems like the optics aren't sitting all the way down now. I'm not sure but suspect that. I did do some light sanding of the shelf and backside of the MCPCB to smooth things out but wouldn't think it would make much difference in thickness, but maybe it did. Really weird, but they had written 5K with a thick black marker on the backside of the MCPCB. The sanding of course took it off.
Ohh - the solder paste used is probably lead free - think it has a higher melt temp than regular leaded so be sure to apply more heat. It's always risky re-using an MCPCB because of the solder remnants left behind, but I think mine came out well. I added more solder but tried to keep it as small as I could.
Glad that worked out for you Tom E. Thank you for that rundown of all the obstacles and concerns, it was very informative on what to expect. Nice skills there I must say, indeed sounds fairly tricky requiring a really steady soldering hand and I think 7 LEDs is also a fairly high # to reflow at once. I can give this info to my light building partner and see what he thinks about getting mine swapped for 3500K CRI 95+. Well, congrats on the successful mods, youâre a gentlemen and a scholar! 
My MF01 mini review is out and it contains the copper heatsink thing too.
Thanks for the video Zozz, good Job :+1:
Just watched on YouTube and commented, thumbs upped, and subscribed! Zozz bringinâ all sorts of flashlight goodness to the dinner table; food for my LED thoughts! 
I want 1.
Did some testing at home, ceiling bounce lux readings.
I have a black SST20 5000K version (yes it is green in the low modes, but nice tint about half ramp up and above).
Used Arctic MX-4 to install the copper heatsink.
Battery was a fully charged iJOY INR 26650, 4200mAh.
Lux meter was a Uni-T 383.
Turbo mode was the same as stock (as expected), and I forgot to test high ramp stock.
TURBO
| TIME (s) | STD (%) | COPPER (%) |
| 0 | 100 | 100 |
| 10 | 96 | 97 |
| 20 | 88 | 93 |
| 30 | 82 | 88 |
| 40 | 34 | 40 |
| 50 | 21 | 22 |
| 50 | 14 | 14 |
HIGHEST RAMP
| TIME (s) | COPPER (%) |
| 00 | 100 |
| 15 | 99 |
| 30 | 72 |
| 45 | 75 |
| 60 | 83 |
| 1:15 | 88 |
| 1:30 | 90 |
| 1:45 | 92 |
| 2:00 | 92 |
My readings doesn't show a fully regulated highest ramp, is it because everything is stock (including thermal config)?
Or did I mess up applying the cooling paste?
Thank you man of light for this mod, and the instructions in the first post!
Yes, probably because thermal config is stock. Thatâs a good thing to check Iâd say. From what I know, default highest ramp should be fully regulated and as expected, but maybe because you have the stock thermal limit still set it reduces output. Try bumping it up, even as high as 85c to test it, and if you want lower from there if you choose (in steps if you like to see what is best).
Iâm using the Molicel at the moment and on turbo it seems to last me 25 seconds before any noticeable step down and its super bright, even though it is in 4000K 95 CRI.} :smiling_imp:
If you suspect your battery may not be putting out enough power or is of unknown quality, it wonât hurt to do suggestion #3.
1) Try increasing the thermal limit and see if it affects turbo and max ramp (my temp setting was increased to almost to no limit or very little limit, I forget exactly, but maybe 85c)
2) Consider what kind of thermal grease, paste, or compound you used (I used Gelid GC Extreme which is a cutting edge, top-shelf CPU thermal compound), and if think you did not put enough, you may need to redo with a tiny bit more, just enough to cover the areas mentioned in the instructional photo and also in the corners where the host meets the outer heat-sink ring.
3) For 21700 battery size, youâll get better performance than anything out right now in the 26650 size as of now as far as I know. I can recommend Molicel P42A (21700), Samsung 30t (21700), Samsung 40t (21700) and Golisi IMR 4300mAh Gold (26550) if you prefer that size (it is almost as good as the 21700, just not quite, but probably still the most hardest hitting & powerful 26650 in FET mode (and likely even in regulated modes). They all seem to be very competitive, just depending on your preference and situation. You could then test it again and see if this helps.
If anyone can try this to duplicate my results for the turbo test, that would be great.
Got mine today, but there is a problemâŚmy light has not arrived yet! 
Haha, same here. I ordered the MT07 on Jan 14th and hoped it would get out before the Chinese New Year but it did not. And with the outbreak problems going on over there, who knows when it will get here. Oh well, my flashlight is of no concern compared to their health.
Yes, health comes first before anything elseâŚI hope this virus thing miraculously disappear fast!
Slight off topic, but that talk about the hardest hitting battery got me wondering⌠I have a Samsung 40t 21700, is that conducive to the best output?
In a nutshell yes, LOL (changed from ânoâ after my good friends here made some good points and explained a couple things to me in detail, and probably because I was blindly biased to Molicel because I have one LOL, how terrible is that⌠I have to apologize and I do take this seriously; I stand corrected in that the Molicel P42A is not the King ( 
) and Samsung 30t & 40t are very competitive and the 40t is conducive to best output, its just a matter of preference on some slight differences in performance, under different conditions.
If you can do a test before and/or after using your 40t, this would be helpful in figuring out if its conducive to best output scenarios; I would like someone to compare results with when generating the best output possible using the three batteries I mentioned. Namely, how long on Turbo before any noticeable step-down. Thanks, and good luck.
Thermal configuration is very important on every Anduril based light, the stock config is mostly not accurate. Temp reading must be set correctly and max. temp. setting should be set to a I reasonable value, I use 60°C (+30 clicks) as max. temp.
Then, if you still have a step down effect in full ramp mode, check if the AMCâs 7135 have a good thermal connection to the heatsink, and whether there is sufficient thermal paste on all marked surfaces.
If everything is correct, you should get a better result.
Oh, in the morning I got my copper plate and lo and behold, my Mateminco MT07 arrived in the afternoon.
Ok, I am no modderâŚ.I have no thermal paste, never had it.
Can I just install the sticker, use no thermal paste, and be done with it?
The Molicel P42A is a 21700 and works with the included adapter which I added some foil around to tighten it and provide some extra heat sinking.
Anduril only goes up to 70°C. If you want any higher you have to set the ambient temp lower than it actually is. If you set it 15° cooler and max to 70° then it should treat 70°as if it were 85°.
Molicel P42A is good, but not the hardest hitting. 30T clearly will beat it at 20A. The 30T is optimized for high current at the expense of capacity.
Yes, itâs fine. One of the best 21700.