MF01 Mini / MT07- Mod/Upgrade - Copper Heatsink Cover Plate

First test happened.
Short result: it works

More detailed:
The edge thickness is a bit high. It leaves a gap between head and tube but around 0.5mm and O-ring perfectly 8nside head so not much deal. I will polish the edge which contacts the driver but without that I see a big improvement. First I made a reset on anduril to try as it is out of the box. Stock ramp max was 2200 lumens. With plastic it dropped instant to around 8-900 and creeped back up and then down… With copper thing it has around 20% drop from 2200 lumens then went up to 2000-2100 and stayed there for 3 and half minutes. After that it slowly reduced output to 750 lumens at 7 minutes. The hottest temp under this test was 43°C on head. I will recalibrate temperature and set 50°C limit for stepdown and see what I got. Now 4 pieces of liitokala 26650-s are in charger to start all test runs with full battery. I made a video of installing and I will add runtime graphs to it and upload that this weekend or early next week.

Thanks for your first impressions.

Please tell us which version of the MF01 Mini / MT07 you tested. Due to your output levels, I guess the 5000K or the 6500K.
.

Well, this is kind of a “It’s not a bug, its a feature” thing :wink:
A small gap between head and tube is necessary to ensure a proper contact of the battery tube to the heatsink plate. In stock case, there is a gap between the plastic cap and the battery tube while it’s completely screwed in.
.

Do you used Thermal compound for this first test?

Mine is MF01 mini 5000K.
Yes I used thermal grease. Can’t tell the exact type. We use at work for industrial electronics heat sinking. It is working well in my dedomed Imalent MS18 so it will do the job. I show full assembly with applying thermal grease in video.

First test was with an Imalent wrapped don’t know what originally made battery. Made a few other tests. With Liitokala it has 2450 lumens stock ramp max. It made the drop at half minute to around 1200 lumens but after that it ramped up and holded 90% of ramp max for 7 minute from the start. I tried raise stepdown temperature to 50 and 60°C. And lowered ramp max to 1650 and then 1367 lumens. That mean in ramp max setting 30 and 35 clicks. The first drop stayed and then hold the output well and interestingly started drop output again at 7 minutes. So I pretty much getting same graph shapes just the height is the difference. I try to replace thermal grease with thermal silicon pads and polish the copper bottom.

I starting to think that the first drop is a program caused thing. It senses a bit fast temp change and scared and starting to drop output to avoid heat up.

Very interesting test results. Looking forward to more :smiley: Maybe Toykeeper can chime in on the 7 minute step down.

Thanks for your hard work! :beer:

Thanks for your work and testings Zozz. Looking forward to your graphs and Videos.

The fact that you have this first drop, at around 30 sec, in all cases, no matter which output level, sounds to me like you still have an thermal conductivity bottleneck.
From my experiences while testing, this can happen if the contact between copper heatsink plate and the surfaces at the driver pcb and the head inner diameter is thermally bad contacted.
You probably already have done this, but please make sure that thermal compound is putted on all surfaces I marked green in my description, not only at the top surfaces of the AMC’s. This can make the difference to avoid the first step down.

I am looking forward to your results, thanks again.

You are right! I only putted thermal grease on 7135s top. I will disassemble and add to the rest of the surfaces where needed.

i know this could change cost, but whats the issue with using a thermal pad?
that way it can be more precisely cut, applied, and with the precut parts everyone knows where it goes. also you can engineer the copper with an exact spacing to create 0 gap.

Wonder how the alu + copper version result vs stock for mf01 mini, but I’m about to order those…

ZozzV6, you’ve got stock right??

Yes I got a stock full aluminum body MF01 mini.

Removed the copper and found the thermal paste barely spreaded out. It seems helped a bit but I think there is much to improve. The gap was too big above the 7135 tops for a grease.
So I tried thermal silicone blocks this time. Cut them to size and the copper thing spreaded them out.

Then I cutted very thin layers to press thm between head side and copper spacer side:

These two graphs with 35 click ramp max setting so 1367 lumens at start. The left was just thermal grease on 7135 tops. The right was the silicone stucked version test same settings. Much smaller drop in beggining and no drop after 9 minutes. Needed to interrupt test because we had a coffee break at work :smiley:

There is nothing between driver ground edge and copper spacer so maybe there is a tiny bit to improve that first little drop but it is much better.
Further tests coming weekend.

1 Thank

I hope this will be corrected by the time of shipping....^^^^

Corrected what?

The first run was not so good because I didn’t followed instructions but better than plastic. Second is almost perfect but I still doesn’t put everywher thermal halping material where it needed. So the copper thing is good just need to follow the instructions precisely. It is a bit hard to put thermal compound to the copper side but it is doable.

Ah ok. I thought it’s design. That sounds nice.

Thx for clarifying, wasn’t paying attention.

@Zozz, thanks for explanation and the latest results :+1:

The idea to use thermal silicone pad is not bad, nice pics, thanks for the work.

It’s really important to spread the thermal compound on the defined surfaces, like shown in the description in post #1 and on this pic below.
Before mounting it should look like this, but here is a bit too much thermal compound paste spread on in this pic, just to show where it should be.

.
After mounting, use a cotton pad to clean up excess residues.

This gap is around 0,1 and 0,2 mm after mounting the copper heatsink plate. It depends on manufacturing tolerances and tolerances of the components (7135) and also of the pcb and soldering process. Designing this gap closer is possible, but it brings some risks. It may be possible that another production batch of driver pcb’s will not fit or there will be no gap. Tolerances differs often on pcb’s.

Therefore the idea to use thermal silicone pad is not bad, but in most cases you will achieve better thermal conductivity with grease/paste.

As you said you can’t design it to get too close to 7135 top because if it has a bit more solder under you can possibly break them tightening the copper. You need a bit of clearance. That is why I tried silicone. It works like a gum and reach the 7135 sides too because of the pressure. So more contact surface to get off the heat. I maybe only use paste on that contact ring on driver. I think I will disassemble it a few times to get best results.

I’ve always learned that thermal grease is only to fill in the tiniest of pores and scratches between two metals. You guys are killing me with just glooping it all over and trying to fill in these giant gaps. :confounded: Lol

Recently, I needed to buy Arctic Thermal Pad and it works quite well in one flashlight. Cutting the right pieces is probably a better solution than dirtying everything around with paste.

Ah too bad I missed this. Would love to improve my 3 mf01 minis!

In for 1-2pc if there is a 2nd batch

I will put you on the list for a possible second batch, will provide it soon in post #2, if there will be enough interests it could be realized.