Does anyone have a TR-J20 yet?

[LEDMatt]

@A number 1 - Very interesting. So all I need to do is add a resister between the contacts on R12 huh? I want to experiment with this.

Tweaking this between 5 and 6 Amps is what I desire. I assume the current increases as the resister value decreases?

Any reports on runtime / duty cycles performed thus far with your mod?

[A_number_1]

@LEDMatt

@A number 1 - Very interesting. So all I need to do is add a resister between the contacts on R12 huh? I want to experiment with this.

Tweaking this between 5 and 6 Amps is what I desire. I assume the current increases as the resister value decreases?

Any reports on runtime / duty cycles performed thus far with your mod?

It’s simple, but not THAT simple.
R12 is the missing resistor in a 4 resistor parallel circuit.
When I measured the circuit (without R12) I got about .50 ohms.
When I used the 26650’s I too got about 3.4 amps tail cap current.
So the .47 ohm resistor just about cuts the circuit resistance in half.
(it’s pretty hard to accurately measure less than 1 ohm, your meter’s leads are about .10 ohm.)
See this for formula to calculate the circuits NEW R value:

Series and parallel circuits

This allows about twice the current to flow.
When I changed the batteries to 32650’s, I now get about 7.4 amps at the tail cap.
I wouldn’t try for much more current that that. I’m not sure the other components will survive.

I didn’t need a LUX meter to see how much brighter it was now than stock, it turns night into day, on my street.

I did notice that the massive head and cooling fins dissipate heat extremely well, very little heating effect.
It’s not made to put on your keychain, but If you want to build inside the pill, there is plenty of room.

I havn’t done any long runs and runtime testing, I have to travel and will not be back for a week.
If you can do measurements - that would be interesting. Good luck.

[lunatictr]

@A number 1

The TR-j20 has two banks of six leds are driven at about 18V. So with about 7amps split between two banks you get, after losses, about 3 amps per led. Don’t know if I would want to try for any more, but It’s quite bright, sorry I don’t have beam shots, I’m too busy now, will take some in future. Good Luck, please post your results.

Unfornately it is not that simple. You measure the amps at 3*4.2=12.6V withouth the sag. but it is converted to 18V and lower amps. You draw about 92W at the tail. without losses you divide this to 12 leds and get about 7.6W per led.Include some losses at the driver, say 10-15%, you get about 2/3rds of 10W. about 2 A per led.

[LEDMatt]

I modified as suggested but with a 1 Ohm 1 Watt resister (which I had on hand; pulled from an LCD ballast board).

Results
Tail Current unmodified: 3.18A (estimated 40 Watts consumption)
Tail Current modified: 4.20A (estimated 52 Watts consumption)
Percent Increase: 32%

Ceiling Bounce unmodified: 187LUX
Ceiling Bounce modified: 235LUX
Percent Increase: 26%

A conservative mod with a noticeable change. I may go for more after some 32650s come in.

Coincidentally enough, I found a .47 Ohm resister on another junk board in the basement! I do have concerns with PCB heat and what would happen in low cell voltage conditions; however, I saw some other great ideas on the forum about cutting out heatsinks for the fets.

At least now it’s in more of an expected range given the size of this monster. Pleased with the results.

Thanks again for the clear instruction, A number 1. Hopefully I can find a resister value to boost it to 50% brightness / current intake.

[B42]

Think it could take an increase to 3 A per LED and use a resistor with even less resistance? Seems like with a sturdy build like that light has it would be no problem at all, though you may need to keep it on medium most of the time after (of course I usually keep this type of light on medium anyways typically)