Modding the driver circuit in "single TR-J18 thrower", Uniquefire UF-X6S...

For the sake of easy reference, I made made this thread for the driver found in various lights.

Uniquefire UF-X6s has it.

Single XM-L J18 lookalike thrower has it.

As with all budget lights, variations could happen in terms of small differences on the driver, and which driver comes with the lights.

Stock output on high seems to be 1,8-2A to the emitter. I had one that did 2A and one that did 1,8A+ (a bit higher output with lower input voltage)

It can be used with 1 or 2 cells. Input amps is basically the same as output amps (with two batteries double the tailcap amps). Its a 5 mode with next mode memory.

Driver diameter is about 21,85mm,

Driver height: 6,75mm

Driver circuit should look like this:

Blue circle shows one of the sense resistors. Based on my very limited testing. The driver does not handle to be pushed to the "extreme" ( 4A+) smelled burnt very quickly. It might be able to do 3A nicely. Adding one R330 will probably give you around 3A. As always, do this at your own risk.

(I saw 2,27+A with R500 added, or 3,8A with R120, but my driver seemed to be 1,83A with normal input voltage from two cells. I would not recommend R120 or such low values)

Red circle shows a resistor valued "0". Remove it if you want the make the driver become a 3 mode with high, "mediumlow", stobe.

(with my slightly modified driver, added R500, I saw 2,27 on high and 0,5x on "mediumlow")

In the green circle is a capacitor Add a low resistance resistor on top of it to remove next mode memory. Then it starts on high every time. (I used a 22K resistor from this package, labeled 273.)


Personally I have no intentions of using this driver in any high/extreme output thrower, but it might be nice for lights where you want 2A output or a bit more. Removing next mode memory makes it much more likeable IMO. :)

Thanks for the info RaceR86. I know it wont work for me but I will have to have a look at what I have and possibly screw up a driver.

I’m waiting for that light to arrive. I guess I need to start looking for a replacement driver. Thanks for the specifics on this one RaceR86.

Great info and picture. Thank you

I stacked a bunch of resistors on mine and had ~2.6 amp tail cap current and no signs of burning.

I have not done any long term run time tests though. I did coat the driver in epoxy mixed with silicon carbide for a thermal potting after I was happy with the output. Also have a cap shorted to start on high after a few seconds of being used, no more cycling through blinkies.

I measured about 70% efficiency…maybe a bit less.

Great info once again!! I think I got the same driver in a Small Sun T16.

ImA4Wheelr linked me over here from pinkpanda3310’s Patrol King thread.

I figured I’d resurect the thread to point out that this driver looks like another QX9920 based driver. The modes chip seems to be detailed here.

There are several QX9920 drivers currently receiving fairly detailed examinations. HX-1175b and the “16MM 8.4V 5-Mode 3A LED circuit board For 2*Cree XM-L” from LightMalls

On this driver the inductor is pretty puny looking and high-value (I assume that’s 100uH? In that case the saturation current must be tiny…) but the diode is already a big one (5A). I wonder how much current it can handle with a simple FET swap?

Sounds like a great idea as this driver is in several lights. Do you have a particular FET in mind? If I don't have already it, I will order it to try it out and report back.

i think mine is a 2-board driver construction. not sure if that’s from my uniquefire or from my longfire, or whatever the name of the flashlight is.
the driver didn’t work anyway, so i’ve been passively thinking about getting a replacement driver. but haven’t bought anything yet…

It depends on whether the stock FETs are P-channel or N-channel. Hopefully it’s N-channel. In that case I’d just use whatever they are using on those small BLF-DD boards now, that should be sufficient. I expect that there are two FETs which need to be replaced, we’ve seen several drivers with that configuration. I don’t really know what the FET that isn’t a normal part of the buck circuit is for, but since power goes through it it needs to get replaced or it will be a problem.