TrustFire TR-J12 5*CREE XML T6 from DX MF and KD

I don't like idea of objects which can make a shortcircuit inside flashlight, can;t you weld some material on cell plus to make it higher? Of course many people use magnets and it works, but it is not safiest method in my opinion :)

Here is lux test with 26650s instead of 18650s compared to 2,8A XM-L C8 and 11w 950lumen desk light:

Tail cap readings are with (4,23v) 3x2,1A and with 2cells 3,3A.

Ok, I got my TR-J12 today.

King Kongs do not work in 3x 26650 as-is. These cells have a bit of a recessed positive contact and in order to make a connection between the cells enough force has to be applied to squeeze them together. You can fix this in a number of ways, but AS-IS they don't work. I have not tried 2x26650 yet.

I have some IMR BatterySpace 26650s from a while back and they DO work in 3x26650. these cells have a bit of a raised button on the positive end and it enough to make contact. However, connection can easily be easily interrupted with a slight bump. (The good news is that it takes more than a bump to change modes.)

So unprotected cells will most likely require adjustment to work properly. I'll report back later.

EDIT: The positive contact inside the flashlight is either flush with or recessed in that white collar you can see in the pictures, so no matter how much pressure you put on the cells it's not going to make contact. The head either need a blob of solder, a magnet (at your own risk),or a heavy spring. Even then, with the KK cells you are most likely going to need solder blobs on the positive ends or magnets between the cells. On mine, the force required to make a connection between cells is so great I don't think it could be done reasonably. (Think old D Maglite spring levels of force).


3x26650 IMR- 2.02A

3x26650 A123 (3.2v) cells - 2.4A

Nice! I cant wait for your TK70 comparison ! (hint, hint) haha

ah wait! Its ergotelis who's doing that! oops! :D

Still nothing received from KD though...Ahhh cHristmas times make shipping time quite high....

Im using magnets on my plus side, what risk am I taking buoying doing this?

Just a couple more thoughts/observations:

Nice build quality


One cell operation - it looks like it goes DD (which I think is what we expected) - 4A on a single KK 26650, 1.6A on an A123 (3.2v) cell

Heatsinking seems good, but I haven't REALLY put the screws to it yet.

Could you please test if 2 x 26650 works?

In general? Yes. 2 x King Kongs out of the package? No.

The positive "button" on the KK is the problem. It actually sits lower than the shrinkwrap on the cell.


I just took out the light meter and got 24.9kcd (249 lux measured @ 10m).

(the cells were not fully charged and the light had been on for about 30 sec.)

tip for extending batteries, for those that don't want to solder, and have magnets

wipe anode and magnet w/ acetone or alcohol, put a thin layer of this on anode, then the magnet and let sit overnight (since it isn't exposed to air).

it is basically conductive adhesive but in a much nicer dispenser than a jar. it costs half as much at radioshack and you could do a few hundred batteries. Store it tip down or the tip dries and clogs.

I guess i'll just solder mine then, seems like the easiest way to achieve a reliable solution..

Still haven't bought a light to use them however. I ordered some king kongs in advance because i was so stocked about the tr-j12 but still haven't pulled the plug...

Bugger about the king kongs

I attacked the positive contact tonight. It turns out that under that nylon retainer with the brass contact in the center is a spring. I took that brass plug out of the nylon retainer and replaced the spring underneath with a longer/larger one. I also drilled out the retainer to make the hole large enough for the new spring to come through. The spring I used is a standard spring from a Solarforce p60 switch or the like.

Connection is greatly improved. King Kongs still didn't work, but I added a solder blob to all of them and now I'm ready to rock and roll.

I'm only seeing a 5% 4% drop in lux from 10 seconds to 10 minutes. :) The heat seems to stay in the head, the body stays pretty cool. That would generally worry me, but looking at the numbers (I know its only 10 minutes), it's looking good.

Did you take a peek at the driver?

I did not. I didn’t want to press my luck on the first day. Functionally, the driver is VERY similar to the one in the TR-3T6, just with more power.

For you and many on this site, it is highly unlikely. For me and some others, we usually do discharge capacity tests (on a hobby charger) on the cells so as to match them for multi-cell operation, in addition to voltage checks everytime we pop them cells into the light. And remember, you choose/alternate between 2 or 3 cell operation, so you need to check. 2 cells might be at 3.9V simply because you used them for 10 mins before, and the 3rd might be 4.09V.

The thing is, folks are selling such lights on Ebay with grey trustfires, blue ultrafires, and some even noname blue Chinese LiCo. One cell might have 1500mAh worth of juice, another just 900mAh. How not to expect an explosion?

2 cells are like 3.3A? That is no joke.

Google reverse charging too....

You can prevent a lot of incidents with just a protection circuit.

But not to worry, it is highly unlikely, but every multi-cell light esp 3 cells, i feel there must be a warning.

I set out to do a runtime test, but I just got too nervous 20 minutes in. It had dropped about 8% but the head was very hot. I just had it tailstanding with no extra cooling. At the 20 minute mark I started using a hairdryer with no heat to blow on it for a minute or so out of every five. I could get the output to get back about half of that 8%. After a blowdrying session, i tried holding the light in hand for 5 minutes and the output only dropped a fraction of a percent.

The additional cooling was evident in the output within 10 seconds, leading me to believe the emitters are very well heatsinked and the head has a good thermal path.

I could use a fan for the runtime test, but I don't know about you, but I don't have an EDC fan (or hairdryer) and I'm surely not going to hold the light for 2 hours (EDIT: 1:30 - the cells wouldn't last 2 hours), so I'm going to assume it will run for that long provided I can keep it reasonably cool.

All things considered thats pretty impressive. I wish this one had a "turbo" mode like the DRY as it seems it would handle the heat much better. On second thought screw the "turbo", I would be happy for 3A per emitter.