OK If I get the news, I will let you know.
I contacted Mateminko and asked if they plan to put this copper washer on new lights (new batch).
I received an answer from them:“We are thinking about it now”
And then I wrote to them:“After you make the final decision, please let me know.”
Why not let Man of Light/Lexel cooperate with Mateminco?
Well, Lexel has already talked to Mateminco while the early development and prototype testing phase of the MT07 about this topic. It seems like they don’t react on this.
Actual Lexel has already sent the technical drawing and the 3D-CAD model with the results of the latest version to Mateminco. We will see how they will decide on this.
If they want to realize such an update, the manufacturing costs will be higher and most likely they will have to set the price higher for the light.
And then they will ask themselves whether it can generate more sales with such an update then before. I doubt it a bit because most customers/buyers don’t even realize the problem or pay attention to it.
I am curious how they will decide on this.
I wished they’d rather implemented a aluminium washer to address this issue instead of going for fan-service in the form of USB-C charging which also has a very negative impact on the waterproofing.
They might in the future or in next larger batches. Who knows. They might implement it in future light designs of other models if people demand it enough. The truth is people demanded USB charging. To a lot of customers it provides many more pros than cons. You can still get some waterproofing against splash and rain while not having to pack separate charger if you travel with the light. It also increases the “giftability” of the light to people who dont need to learn so many things about LiPos (just basics) and it also increases their possible customer base. They can sell combo packs with cells (which they do) to people who wouldnt buy them if they didnt have charges implemented.
I really, really, like the C charging: I have mine wired up to a belt mounted battery bank TOMO D2/A2 in a small walkie-talkie tac pouch at all times (msg me if you want to know the best pouch). I now get max turbo (max charge) burst basically all the time now everyday/everynight. Always @ 100% charge now! An extra 12000mah 2x 26650 vapcell (5900mah) added to my 4000mah Molicel hard hitting 45A 21700 (higher lumens) battery in the light. Yeah its beasty, and with the heat sink mod it will be just splendid… splendid I say!
So, I have now ordered an MT07 Xpl, I have already ordered the copper plate from man of light.
But what about heat dissipation from the LED board.
I also like to swap the original rather poor thermal paste for more effective ones from eg MX4.
Can you raise the LED board so far to wipe away the old ones or apply new ones, or are the cables too short?
I mean the circuit board where the LEDs are resolved, it brings a lot, I do with everyone where you can get to it!
Not sure the heat dissipation from the LEDs is bad. The light head and the entire light body gets really worm really fast on turbo. I have seen some footage of the thermal paste between the leds and aluminium body and there is some. Not sure what type and quality and you sure can swap it for something better but I would not expect a performance boost. The light seems to absorb the heat from the LEDs as well as one would expect. That being said, you might theoretically increase the longevity of the leds by better connecting them thermaly to the light body. But you will feel better with having a better thermal paste there 
(I might change it too once I install the mod). But it doesnt look like an issue compared to the AMC thermal thing.
There may be a similar issue on the Astrolux FT03S, but this time it’s the FET that might be heating the MCU, not the 7135 chips.
Really? What is the gate voltage on that one and what is the drain/sourcu current? Thats very interesting. Usually they produce very little heat. Do you know what MOSFET is there? Did they use logic level mosfet?
I don’t know what FET is being used. There are very few pictures and the driver has a charging pcb right on top so it’s hard to see under/around it.
We are trying to figure out why this FT03S seems to step down after 20 seconds on Turbo and the exterior of the light is just barely warm. At the top of ramp it has no issues, it lasts minutes and the exterior gets nice and hot.
It’s a bit of a toss up if it’s just software reacting too quickly, it uses Anduril, or if maybe the FET is heating the MCU prematurely.
I was thinking maybe a chunk of copper could be attached with thermal adhesive to the FET to slow it’s heat build up, much like with this MF01 mini, and see if it allows for longer Turbo times. I don’t own this light so I can’t experiment myself. I thought I’d mention it here to see if any of you technically mined folks had any thoughts on the situation.
BTW, the FT03 uses a different 2 channel FET+1 driver with NarsilM and it’s xhp50.2 version draws almost as many amps, yet it runs fine on Turbo. I know the software has to play a certain role in the FT03S stepping down early, but is it all due to software? IDK.
Maybe there is a software limiting function on turbo (but I don’t think that is the case). Someone on banggood questions asked why these have no thermal compoud so there is a possibility that at turbo the heat transfer is delayed and they get hot? In that case a thermal paste could fix that. I think if would be a good idea to ask Texas_Ace for his opinion. I think he designed the texas avenger drivers in those things. Could be a fet heating up problem, even mosfets can start heating up above 10A and who knows what internal resistance that fet has at turbo gate voltage. Maybe the driver is pumping more current on turbo into SBT90.2 than into XHP50.2? I have more questions than answears. One on the users that has thamshould try contacting Texas_Ace.
The first several FT03S showed up with no thermal paste and one was missing a lens. 
It’s like they rushed them out the door. Current ones do use thermal paste and the people that report the premature stepdown on Turbo also have thermal paste, so I don’t think it’s related.
The Anduril software has a pretty complicated thermal code. It tries to anticipate when the set temp will be reached and start reducing power early so that the light won’t overshoot the limit. Normally Toykeeper will rewrite the thermal code for each specific light if shes involved in it’s design. Both the FT03S and MF01 mini used existing versions of Anduril. Astrolux did not consult Toykeeper in either light.
The MF01 mini thermal code seems to function quite well once the heat from the 7135 chips are better controlled. It’s possible Astrolux used the exact same version of Anduril on both lights since they are both so similar in design.
I’m not very familiar with the nuances of FETs. I know their internal resistance can vary based on the exact model used. So certain models can limit output or boost output when used on FET drivers. I’ve heard of some FETs burning up due to too much amperage. Since the FT03S can pull 20A I assume the FET might be getting really hot and effecting the MCU temp.
I do notice the FET is right in the middle of the driver as opposed to the edge. There must also be heavy traces between it and the led wires. The led wires are right next to the MCU. It seems like a decent heat path.
I wish I had a light to test on. I’m just working on this problem because it’s interesting.
In Toykeeper’s Anduril firmware repository here:
http://toykeeper.net/torches/fsm/?C=M;O=D
The MF01-Mini (MT07) has an Anduril build target for it, so it’s likely optimized.
The FT03S and EC01 do not though (the Anduril firmware used in the EC01 appears to be an older version that doesn’t have the “manual/automatic memory” toggle).
The FT03S Anduril version has ‘factory reset’ function, so it’s fairly recent build, maybe it’s using the MF01-Mini’s Anduril version?
It’s funny, I specifically asked TK if she did the optimization of Anduril for the MF01 mini and she kinda skipped answering it. I may have assumed she didn’t. Seeing there is a specific version in the repository must mean she did optimize it. I’ll edit my earlier statement.
The FT03S might be using the MF01 mini version of Anduril. IDK, but it seems likely.
I'm working with the cfg and hwdef files for the MF01 mini now. They clearly are marked for the bank to have 7x7135's but it actually has 6x7135. Not sure if it makes a difference. Nothing easy to change - the ramping table would probably have to be redone if it was generated for a 7x7135 config. I'm leaving it as-is for now. I tweaked the voltage reading for my mini, and eliminated the "good night" and beacon modes so the temp mode immediately follows the voltage reading. Also I'm allowing 90C max.
She also has MF01S config files.
The stock Anduril config on the mini doesn't light up the switch LED's when on and ramping. I enabled that - I like having the switch LED on when the light is on, plus seeing where it switches from the 7135's to the FET (switch LED's get brighter on the FET). Downside is the AUX LED board also is enabled with the switch - maybe they didn't want the AUX LED board on when the main LED's are on?
I'm not sure that's a problem - amp draw should be pretty low.
I got two mini's: an alum one I bought early on, and recent purchased brass. I found the alum one has an optics problem - the optic is not sitting all the way down, resulting in a clear and obvious difference in the beam. The brass one has a well formed center hot spot while the alum one has a wider fuzzy defined one. This results in a much lower throw # in the alum one. I did some swapping and measurements - it's not the optic, but clearly the alum one has about a 0.2 mm, maybe more, lower profile and you can feel it - the optic kind of floats, not a tight fit at all.
Anyone else notice this? I took pics of the two and you can see the differences in the top view down on the LED's.
If I got time, my best option would be to add a thin copper shim under the MCPCB, then I'd have to change out the LED wires - tricky in this light. Plus risky - may crush the LED's if not done precisely, I'd suspect, based on other issues of height differences between SST20's vs. XPL HI's.
Fake dessicant in box. Whoever their supplier is, is sticking sand in the dessicant packs instead if the labeled ‘silicagel’ scuffed up the finish on my recently received mf01 mini Copper pretty good. Was barely punctured in my Aluminum version, so no scuffs on that one. I’d rather them not include it. Buffed out the copper and put real dessicant packs in the box. Copper doesn’t look coated.
I will take a look at the lens and LED gaps while I have them open to insert the copper I got from man of light. The beam pattern on both of mine looks the same. Except one is 4000k 95cri and the other is 6000k xpl-hi.
Tom E - do you mind posting the photo of your beam shots?
I don't have beamshots but in these pics, you can see the difference. Alum (green) on left, brass on right:
Notice the gray colored, blobs around the LEDs? The brass one has some of the same effect but much smaller. This seems to be caused on the optic not being fully seated. Here's another pair of pics:
Stock MCPCB/LED's:
What's this? Oh, it's a 5K (5000K) SST20 board. This pic was taken after cleaning with isopropyl - it originally was darker magic marker writing :FACEPALM: . It came off easy enough with sanding.
To me it looks, and measures like the alum one has the optic sitting higher above the LEDs than the brass version, but the top of the optic is dead even between the two:
