BLF style 18650 1s thrower( poor mans Olight M2X)

Wow, the butcher job is actually not bad at all! Like it!

Just drilled some more holes in the pill and DTP board…
Board is now fixed with tiny screws. (flush / counter sunk)
I’m making a habit of doing that since i shorted my MT-G2 C8 because the PCB cut the wires when i screwed the pill tight to the reflector…
In this light the same could happen.

Jerommel,

You know when you said that you’ll not buy 5$ 26mm mcpcb? Well it seems like you spend more than 5$ of electricity for that mod :smiley:
But very interesting indeed but it will not be better than simple 26mm mcpcb mod(better heat transfer, larger copper board).
But yes same ugly aluminum pill like mine… Donat hole in the critical place for heat transfer(below emitter where the greatest heat happens).
For proper and successful modding, and because of mentioned bad pill at least one minute of stable output test on lux meter is must. Mine as I said has 260 kcd for first 5 seconds than it starts dropping fast to 150kcd after 30 or so seconds. That of course will change instantly when fat 26mm mc arrive.

I tweaked some things , it’s now making 210kcd with the stock emitter . Of course it could do better , but it is clearly a super cheap flashlight for what it offers . Heatsinking is pretty good , i don’t see any drop caused by the temp of the emitter .

U4 emitter and AR coated lens would help , and even as Mitko mentioned a xp-l dedomed would throw even better.

With 26mm mcpcb?

Yes . It requires quite some time to sand the pill and make a nice flat surface for the mcpcb , but it pays off in the end

The one that will modd with 16 or 20mm pcb should really pay attention on heat transfer. Donat hole in the middle is simply terrible.

Hardly wait for my 4 x 26mm ones + additional 3 x Mitko super throwers.
Also in meantime our member Driver have sent me original good old XP-G2 S4 2B emitters. So I hope that I will touch or surpass 300 kcd mark with them.

The bottle neck of the heat transfer is the LED itself, not the transition from the DTP board to the pill.
The surface area of the DTP board and the pill shelf is many times larger than the surface area of the LED thermal slug.
Sure, the pill shelf surface is not perfectly flat and has a little dent in the middle, but when you make sure the DTP board is fixed tightly to the pill (i use screws) and with a little thermal grease, it will be fine.

I do wish i had a good old XP-G2 S4 2B though…

Sheesh… That was a lot of work and finicking, but it’s finished now and working.
So i have a XP-L HI in it on 12x 7135, so it’s pretty bright.

The worst part turned out to be coming up with a centring gasket…
Had to make my own from a piece of plastic tube, and had to glue it on the LED board, because:

The focussing.
Turns out i get the best result with nothing between reflector and LED board.
So the centring gasket has no bottom flange, so that it can go all the way towards the reflector.
Had i known this, i would have made the centring gasket thinner, because now it sticks out a little bit inside the reflector, slightly shrouding the sides of the LED, but probably not noticeable as a loss in output.

Thankfully the reflector doesn’t touch the + and - traces of the LED board.
I can adjust the pill from the back of the head, that’s why i made 2 slots for tweezers on the bottom of the pill.
I think i’ll put something thin (kapton tape perhaps) in between LED board and reflector for a little piece of mind eventually.

One thing i think sould be improved yet is the thermal path from pill to head, which is only the threads now.
Maybe a ring of some sorts between pill outer ledge and the head.

Dark outside now, i’ll have a look if the beam is as parallel as possible.

LOL Yeah, it was just barely screwed in there, about half a turn or something.
So i decided to recess the driver board a little, which was not easy without a lathe, but it worked out okay.

This light needed a LOT of work…

I was thinking about your comments when i was busy doing this light, and i managed to flatten the pill shelf a little with a rotary tool and a disc with some sanding paper on it.
The little cavity in the middle is still there though…

And maybe you’re right, it would be better and easier to use a big DTP board, although i wonder how it should be mounted on the pill…

Like this BLF style 18650 1s thrower( poor mans Olight M2X) - #205 by giorgoskok

Post 205

Okay, that’s just clamped between pill and reflector.
But it will make the pill even taller / longer, and then you have to cut off like 3 mm form the front end of the tube, and like 2 mm from the rear of the pill.

Here’s the result of my modding:
(click on pics to expand)

I ruined the tiny Philips head screws when i tightened them with force. :person_facepalming:
Not sure if i can get them out again…
I use a 1mm diameter copper wire (tinned) as a ring between head and pill.
That way i can tighten the pill in the head so that the threads make a good thermal path.
Raising the pill 1mm is just about right for a tight focus (fortunately).
This means the pill is not pushing the LED board against the reflector, i estimate there is around 0.3mm space between LED board and reflector.
Everything stays rather cool, i expected more heat.
I couldn’t even find my Fujik thermal goo, so i used a bit of silicone grease for better thermal conductivity (filling the gaps)
Seems to work. :slight_smile:
And silicone grease doesn’t dry out.

About the output:

Compared it with Convoy C8 with XP-L HI U6 and this Supwildfire is a lot brighter and has a considerably tighter beam. (of course)
BUT…. …then i compared it with my Brynite B158 with (one of my 2) dedomed old stock XP-G2 (bin unknown, tint was 3D i think, maybe even 3U) on 8x 7135 and the Brynite leaves them totally in the dust… I mean, totally…
…and then i took out my Jacob A60 with dedomed XP-E2 R5 on 6x 7135 and it blows this Supwildfire away too, and so does my Uniquefire UF-T20 with better (but still a bit crappy) lens and also a dedomed XP-E2 R5 on 6x 7135.

So, i think i should either have used my last good old stock XP-G2 in this light or buy another aspheric zoomer…
Yeah, i expected more…

(edit) I have a rather bright corona around the hotspot.
Maybe my reflector is not as good as others?
I don’t know…
The C8 has a similar beam pattern (but a larger projection), so i guess it’s normal…

Hmm , maybe .

With dedomed xm-l2 focus wasn’t good with 1mm . My spacer is paper thin now , and focus seems to be just about right .

I think without a spacer would be ideal , but i dont have much time to create a custom centering ring now .

You have misunderstood what i wrote. (it was not clear what i wrote)
The 1mm distance is the copper wire ring between pill and head, so it’s under the pill rim / ledge.
The space between LED board and reflector is probably only 0.3 mm.

Yep sorry.

I edited it, it wasn’t clear.

You did fine and very interesting job here with your modd but that is just to much job. For similar I guess I would need couple of hours myself.

In 26mm mod pill will not be taller and you don’t have to cut anything. It is as easy as gluing thin copper disc(have you heard about artic alumina?), file, rasp or even lathe to make everything flat to that copper disc, 26mm pcb up on top of that and you are good to with: either gluing(AA), just thinly thermal paste(reflector has fairly enough pressure to mcpcb surface), or you can even add a screw(which imho is totally unnecessary).

I think you must buy Osram Oslon Black Flat(few of them than choose the best like Djozz said).

And BTW where are lux meter results?

Edit:
“Everything stays rather cool, i expected more heat.”

That is first sign that something is not right.

Same happens with mine current modd (26mm mcpcb are still on boat) but since I have lux meter I see that I have 260 kcd in first 5 seconds and then drastically drop to 150 kcd with good old XP-G2 S3 2B. Current draw around 4.5A.

In my light I got the best focus with both the centering piece paper-thin and the reflector base sanded thinner.

I don’t have a lux meter.

Well, it turns out one of the 12 7135 is not working, i measured 3.88 Amperes.
For the rest, it is a big head, so it just doesn’t get very hot.
The thermal path is good.