Did you adjust the temp settings of both lights prior to testing?
Yes, I did.
No they’re not pads, just paste so thick it behaves like a putty and comes in rolled up strips. Goes by so fast in the video I’ve not been able to grab a screen shot. 
Those thermal pads you linked to on Ali would probably be good for the top of the resistors. Do you know recommended thickness?
Thanks!
Thanks for this. This type of side by side comparison is very helpful to me and those who say that turbo should not be affected by this mod. :laughing: I knew I was right about it helping with the turbo… This only confirms my positive results! In my case I went from 3 seconds to 25 seconds before noticing even the slightest of dimming… and then the step-down slope was a much better curve from what I could tell. :+1:
I might be able to squeeze out up to 1 minute if I raise the temp control to 30 degrees above room temperature according to another tester who received those results!
Did you happen to get the times, or can you run again and post how many seconds you got before any noticeable step-down on each? Thanks in advance!
ps. Can’t thank ‘Man of Light’ enough for producing this heat sink… here’s a very meaningful thank you #2!
I didn’t check the times. I was just interested in whether my first mod attempt worked or not. The first MF01 Mini I installed the copper heatsink plate in was my sand color light. I held the sand MF01 Mini in my right hand and a stock olive green MF01 Mini in my left hand, turned them both on turbo and watched the beams display on a wall. Within a few seconds on turbo, the unaltered MF01 Mini began to diminish its brightness. The modified MF01 Mini grew warmer much faster in my hand, but I didn’t notice any decrease in brightness. If I had to guess, I probably kept both lights on turbo for around 20 seconds.
I used green Shockli 26650 5250 mAh cells in both lights.
I’ve just updated the OP (post #1) and post #2 with some fresh information. The 3rd batch is now available, all the interested were informed via PM.
If somebody want to buy some of them, please feel free to PM me directly or post here in the thread.
Just saw this post before I head off to work. I’ll get the payment sent tonight when I get home. Thanks mol!
Didn't post it yet, but I did rcv my qty 2 of these heatsinks, and installed one in my green alum one that's been modded to SST20 3500K's and the installation went quite easy. I did end up using MX4, even on the 7135's, and confirmed it's making contact with the heatsink. I also added conductive thermal grease to the top of the tube threads. I was hoping to see the heat get to the battery tube quicker but can't notice it much, least by hand. I'm usually running a GOLISI or 40T in it, with a spring bypass, so it needs all the thermal help it can get, and seems to be making a difference with longer full max runs.
Now I have to do similar mods on the brass mini - update Anduril (my version with voltage calibration), improve the thermal contact between the MCOCB and shelf (sanding smooth, MX4, etc.), then add the heat sink.
Something seems wrong with that. My tube gets hot a bit faster than the head (especially the upper part of the tube; fairly quick @ about 30 seconds, and really hot after 1 min, not extremely hot unless I go another 30s+ longer (while bumping turbo again after any step down) with room temp calibrated at 25c, then 60 clicks to set an 85c cap). I mean I noticed this right away as soon as I finished the mod and tested. The head and bottom part of the tube & tail-cap heat up more soon after.
I added extra thermal compound around the outer circle edges (to fill the gap) and up the threads 1 mm or so.
I used this: Amazon.com
It is top performance at about the same price (prices have lowered) and I think is among the best brands and works awesome up to 180C (which is hot enough to fry your hand like bacon JK/LOL)! :disappointed:
There may be some better than this (and MX-4) if you want to pay more; check this article on the best 85 thermal pastes: Best Thermal Paste for CPUs 2024: 90 Pastes Tested, Ranked | Tom's Hardware
Most will do just fine and because of this, I wonder if it is the gap area that needs more attention (not the paste used)…
I also wonder if the green version’s inner dimensions are somehow slightly different, and causing this issue? I doubt it though, but thought I’d throw that out there.
You would and should feel an immediate difference in the battery tube heat, just by hand. It was VERY obvious to me and caught me off guard even. Would rather have that heat at my hand than at the driver chips though, which is the whole point.
One interesting thing in the Tom's Hwde article is the shelf life of thermal grease. Though I've used many types of greases, and many tubes of greases over the years, my electrical&thermal conductive grease is probably 4+ years old. I've used the GELID GC Extreme before. I've been modding lights for 8 years, and owned and built maybe 200+ PC's since 1981 (bought the original IBM PC new - I do mean original...). I've gotten some of my best info from that site in the past, but it really turned heavily commercial in the last several years. Can't recall why I switched from the Gelic GC, but might have been because it's either less viscus or tends to dry out - not sure.
I'll try stripping off the grease on the tube, and use the MX4 instead. I was just concerned about losing some electrical conductivity because that batt tube connection to the head is where the batt- returns to the driver. But Nyogel is not electrically conductive and doesn't seem to effect the batt- connection, in fact I swear I see improvements in amps/voltage after applying it on threaded joints.
After chatting with you guys, I went ahead and purchased a couple of Samsung 30T 21700s to have the hardest hitting battery for turbo as possible.
Here are my updated results using the upgraded heat-spreader mod, first just increasing the thermal regulation cap, and then second using the Samsung 30T (and compared to the Molicel P42A):
Using the suggestion of another member on this thread (about a week or two ago), I tried to duplicate his fantastic results by increasing the temperature cap settings (to even more than what I had set)…
So, the first thing I did was calibrate the room temperature to 25c, and then clicked 60 times to set the thermal regulation to cap out at 85c (up from a 65c).
And using a fresh Molicel P42A 21700 4000mah battery, my length of a cold start Turbo has increased from approximately 24 seconds (without any noticeable step-down) to about 36 seconds, a 50% improvement with just a setting alone!
Next, I compared this to using the newly acquired Samsung 30T 21700 3000mah cell. Well, it seriously made a noticeable difference on the ceiling bounce test which is good enough for me (for the time being; I would love to purchase a lux meter rig in the near future). As expected, the heat went way up in relation to the extra lumens being generated! :exclamation: This thing can get hella toasty now after bumping the turbo button 2 times after each noticeable step-down, which will maintain near full brightness for about 1 minute until the light becomes way too hot to hold and would require gloves to test it any further; I would be scared to push it but it should be able to take it as long as that battery never reaches 85c which it probably wouldn’t (hopefully), even if the driver and host shell was actually 85c, it would take some additional time (and the plastic battery adapter seems to provide insulation to the battery).
So after about 25 seconds with each battery, the Molicel P42A had about 91% juice left according to my battery charger. The Samsung 30T had 89% left and the voltage curve gets worse than the P42A from there on out, but that’s the trade off for that insane amount of light coming through the front of this rather tiny pocket monster. I think next I’ll try to bypass the double springs using 18 gauge wire (if that sounds good to any of you guys). I’d like to hear from Tom E. or anyone else on what your results doing a spring bypass achieved and what gauge of wire did you use? Thanks buddies! :laughing:
Oh, and I forgot to mention… I got about 24/25 seconds of max Turbo using the Samsung 30T (after upping the thermal ceiling to 85c). That seems very decent to me as I was used to that amount of time already with the Molicel P42A before upping the temp ceiling. :+1:
UPDATE 1: So 40 clicks is the actual maximum and will do. I will be using Tom E.’s suggestion below and update and second time with the results later on today/tonight.
UPDATE 2: Unfortunately, I didn’t really see a difference after changing the temp cap and lowering room temp, so I take it I am running it at about as good as its gonna get for now. I’ll test again after a spring bypass and update my results again.
I used 20 AWG - note sure if I recorded results. I use 20 AWG when I can, if it fits, otherwise for small springs, use 22 AWG.
Anduril stock max's out at 70C, so 40 clicks is the max it will do. If you click more than 40 times, it takes the max of 40 clicks. The trick to go higher with the max temp is to set room temp to like 5C, so this gives you another 15C so you can max it out to 85C real, even though Anduril thinks it's 70C (assuming room temp is 20C).
Ahh, thanks for this; this is golden to me, I can probably get even better results knowing this.
I also use 22 gauge for smaller springs, so that seems to be the norm. For the Mini, I will go with your suggestion of 20 gauge wire and bypass it from the top of the big spring with a smooth and even flat cap tinning, down to the bottom of the smaller inner spring (tucked between the bottom spring and tail-cap), actually it looks like there is a brass area in the middle I could solder to. Is that where you went or did you tuck and solder it under the bottom spring?
No worries, in general I am guessing an average of 10% boost in lumens in most bypasses for FET mode, and maybe that or less in increased energy efficiency (5–10)? :question: That sound about right?
Small update and corrections on post #2, interested list.
Thanks MoL! Looking forward to applying the mod. And Thanks for all your hard work to support the community.
… .… — …- . …. .… .- … …. .… … —. …. … - —- —-
Buona sera a tutti mi sono registrato adesso e sono italiano uso un traduttore per scrivere fatemi sapere se a voi traduce bene la mia lingua o bisogna scrivere in inglese
Hello Tony and welcome to BLF.
With the help of Google translator your postings are easy to understand.
Unfortunately, I never learned Italian
Ok ho ordinato su banggood una mf01 mini ora cercando sul web ho visto questo bellissimo forum e ho visto la piastra di rame che avete creato ne vorrei acquistarne una se possibile
Thanks for your interest, I will put you on the List and send you a separate PM message with all important informations.