I purchased my TR-J19 from Fasttech with the intention to mod it following 18sixfifty's great success with his. Not the most powerful triple after modding but for an inexpensive mod this thing is a beast! The best thing of all is I'm into it for well under $100. $)
And what exactly did $92 and some time get me? 205,435 lux @ 1m and 906.5m of throw
Here is the stock setup. The driver appears to be exactly the same as in the TR-3T6 and the J19 only does 1363 OTF lumens stock, just like the TR-3T6. The reflector and pill are massive!
Here is what I swapped the stock emitters for. XM-L2 U2-1A on Noctigons. I dedomed them in gasoline in case you didn't notice. ;)
Rewired them in series using silicon insulated 20awg silver-plated copper strand. Arctic silver was used under the Noctigons. I also chamfered the lip in the centering rings to reduce the slight amount of light blocked by them.
I replaced the wiring from the DRY driver with the same. I stripped the components from the stock driver and used it as a contact plate.
On the switch end I added a piece of the 20awg wiring for less resistance and soldered the heavy brass cap to the end of the spring.
The stock setup for the anode spring is a floating design. The driver retaining ring contains a nylon insert into which the driver spring is set. Tightening down the retaining presses the spring against the back of the driver plate. I have no pictures of what I did to this but simply put, I pulled the spring form the nylon then bored out the nylon using a stepped bit so that the spring moves freely inside the nylon. I could have removed all of the nylon but keeping it retains the insulating property so that the spring cannot short out on the retaining ring. I then simply soldered the spring directly to the back of the driver contact plate. On the battery end of the spring I added a 6mm brass contact button so it makes better contact with the cell top.
Here are the original output specs and the modded specs. I didn't take a lux reading from stock but I can assure you it wasn't that great. My numbers are taken running King Kong unprotected ICR 26650 cells. The DRY driver is DD in Turbo mode. Different cells will result in different output results and I'm sure even better numbers could be attained with even better cells.
I’m only seeing about 3.28A at the tail but that is with very long stock probe wires. I don’t list tail amps because I find them to be utterly useless and inaccurate. Better than nothing I suppose but different people will see different results with the same exact light depending on the thickness and length of their DMM leads. When it comes down to it the only accurate measurement is a good light meter and a calibrated IS or light box.
The cells used are listed in my OP right above the output tables. Unprotected KK ICR 26650 ;)
I normally check emitter amps. That way the switch is included and I get a much more consistent reading by soldering a wire directly to the emitter. It says a lot about if you have reached the max potential. In the lights that are essentially direct drive I think it says a lot. Your amp and lumen number seems to somewhat make sense... But the question is, why isnt turbo higher?
Im not too familiar with the dry driver myself (still waiting for one and never used one), but 3,28A at the tail and 2150 OTF lumens both seem quite low for your direct drive setup with KK ICR cells... Sounds like something is limiting the peak output..
Your SRK lights have around the same OTF lumens, I would expect more from this setup. Seems like there is something that still can be improved...
Yeah but I remember Tom E’s Super Shocker mod that being directly driven at 5~6A to each de-domed LED his Shocker is able to put out 4000~5000 OTF lumens. The driver he uses is different though which is from IOS, but if the DRY driver is able to do DD in turbo mode, which is above 4A to each LED then your light should give at least 3500 OTF lumens, just saying.
It does seem a bit low lumens wise for XML2s, but I’d expect runtime to be awesome with 32650s and very little heat to kill efficiency due to great thermal transfer and massive head.
I just checked mine by eye doing a ceiling bounce up against a Fandy Fire UV-S5 (SRK clone) one of the early good ones that is actually a little brighter than the SRK's I have had and the J-19 is a LOT brighter even though the tint is not as white. It lights up every nook and cranny in the room where the FF doesn't. It's not terribly scientific but I can tell that the J-19 is way brighter. Both with fresh batteries as well. I can't see anything that you did differently so it's strange. Well on this one I did leave one emitter's dome on and stretched the spring a bit to make better contact with the board, it's kinda jammed in under the plastic piece. When my light gauge was still working I tested the one I sold to Gords with it and it was showing about 1,000 lumens more than the FF. Now that could be because of extra light bouncing around but I still would have guessed at more like 3,000 lumens, very close to my terminator. I didn't use the centering ring for the emitter either, the Reflector is sitting right on the noctigon with electrical tape over the wires so it doesn't short, I have found that when I de-dome that it hurts some if the emitter isn't up into the reflector as much as possible. Well anyway it's one heck of a mod and thanks for posting the numbers. Maybe my eyes are playing tricks on me. I have another lux meter on the way so when it shows up I'll have to build a better box for it and post my numbers. Here are pics of the stock J-19 and my newest J-19 mod.
edit, I just had my wife do a "blind" test for me with a ceiling bounce. She didn't get to see what lights I was playing with and she confirmed that the J-19 was far brighter than the FF SRK and that the J-19 modded was also much brighter on High (not turbo) than the J-19 stock was on high. In fact the J-19 modded was slightly brighter on high (not turbo) than the FF SRK although it was close.
One more edit: Just another slight difference I used Wakefield Deltabond epoxy for the emitters. Can't think it could make that much difference though.
I’d be expecting to be much closer to 3000 lumens than 2000 just because an xm-l2 u2 at 3a on alu is rated to be comfortably above 1000 lumens, your direct bonded to copper, yes their dedomed but emitter lumens should still be above 1000 so your losing 800 - 900 lumens somewhere.
On the other side, is it possible that the DRY driver doesn’t deliver REAL high current in DD? As JohnnyMac mention that he measured only 3.28A at the tail, is this where the problem is?
it always has done in the past, I’ve got a couple now (one none step down model). but thats besides the point.
the 3.2a tail current is at 12v, I’m not sure its the same as emitter current, but even if it is, it still comfortably puts those three xm-l2’s over 3a with better than average heat sinking.
How many amps are you seeing on the tail of your light gords1001, and are all/both springs copper braided?
Amps on the tail should be the same as emitter amps. The voltage of the 3 batteries in series are divided on 3 emitters since they are wired in series.
If the driver is proper direct drive on turbo without lots of resistance, the batteries are good, and there is not a lot of resistance somewhere, amps on fresh batteries should have been much higher than 3,3A, and OTF lumens should also be higher.
Sorry JM, there was no intention to take away all the wind from you sails. Just wanted you to have a light that lives up to its potential and what you probably modded it for. :)
Tom-E normally uses this driver since its direct drive when using 3 cells in series and 3 emitters in series.
Maybe try your light without the driver circuit. Test proper direct drive and see what kind of lux and amps you get in comparison. If its the same, I would double check resistance between batteries and the driver..
Edit: Before doing that, do you have another set of batteries you could try it with. Just in case you have a bad cell or something..
I’m not doing tail current tests at the moment, I’ve been getting too many wonky readings and until I figure out why, its not going to be helpful.
I can try with the clamp meter tonight if need be though,
johnny, do all you king kongs drop right through the battery sleeve without interference? And is there definitely some resistance when screwing the tail cap on?
for those who’ve not seen the guts of this light, everything is HUGE I sleeved some 18650 cells into it to test and they wouldn’t reliably run it, the plungers are almost the size of an 18650 so those puny cells don’t work so well lol
are you going to do any beamshots ?
