Today I did this build:
Which is to swap the stock driver with the KD 5 mode one. Reports of 4000+ lumens. We shall see
Heres a stock look at the internals of the TR-J12
Opening the head and tail cap for access was easy actually.
I used calipers and they popped right open!
old driver next to new kd one
The new driver did NOT drop right in, I had to take a rotary tool to both of the boards to slim them down approx 1-2mm before it fit right. No big deal, and I got to use powertools
My soldering skills are not great, but theyll do. The positive contact is a ball of copper braided wire with solder
I then created my own isolation disc.... out of... electrical tape..... this is a budget mod!
modded the tail cap spring (there is braided copper wire covered in solder to take the heat. From what I have heard this thing will melt springs especially when the voltage drops and the driver sucks even more amps from the imrs.
tail cap assembled. (I later put the piece of plastic back in for isolation)
I was definitely nervous about firing this one up. I repeatedly screwed and unscrewed the head in to make sure I wasnt spinning the wires around too much and damaging them. In fact I purposely twisted the head around counter clockwise in the air so that when I screwed it in it actually untwisted the wires.
Still a bit concerned after checking my connections. I definitely did NOT want 3 26650 unprotected IMRs going nuclear at full charge
yes that photo is a bit blurry, and I don't care... my camera sucks - and I blew all my $ on flashlights.
The only bad thing is that I am getting some serious PWM flicker on 5% and 25% modes... sometimes it even happens on 50% Bad form!
Oddly, sometimes if I cycle through modes again it stops I was a bit annoyed at first since other modders did not get the same results I did with PWM. Maybe I will mail KD and ask them about it. But then again, I really did not build this for the 5 25 and 50% modes now did I?
Also, the stock wires in the tr-j12 suck. They're too thin for sure. Perhaps when I get more adventurous I will replace them with teflon wires.
So how bright is this? And how much improvement was there over stock.
Well before we go to the numbers let me just say that THERE ARE NO DISCO MODES. This was a huge thing for me. I have no use for modes that are likely to induce a seizure
Here is some raw data (all measurements were taken at turn on with 3 freshly charged mnke imr 26650s with the flashlight/light meter in the same place under the same conditions)
Ceiling bounce (for added reference my sst90 mag running on 9a dw driver put out 200 lux on this same test)
Low - 42 lux
Medium - 178 lux
High - 288 lux
KD driver TR-J12
5% - 88 lux
25% - 202 lux
50% - 291 lux (essentially the same as stock high mode!)
75% - 384 lux
100% - 491 lux*
* Yes I feel that I must put an * here. This is because I do not believe the 491 to truly be an accurate measurement. I have read reports here that this thing puts out like 80+ watts for the first few seconds on fresh cells and then quicky drops to around ~55 watts (stock driver is 29w)
Needless to say after about 15 seconds that 491 dropped like a brick!!
After about a minute it settled down around 425ish and then slowly started declining about 1 lux every 3-5 seconds (most likely due to the heat this thing was throwing off)
In this test 100-115 lux ceiling bounce seems to equate with close to 1000 lumens. So 4000 lumens does seem accurate.
This was a fun mod and took about an hour. I think it was worth the $20 I spent on the driver. The 100% mode on this is visibly brighter than the dry 3 xml cool white I am thinking I will replace the crummy stock wiring when I get a chance. As for the PWM flicker - I guess I will live with it for now.
I believe this is the cheapest way to achieve 4000+ lumens. I paid $58 for my TR-J12 (it can be had now for 49 from ali express) + $20 for the driver. Cost me $78. The dry goes for around 68 on cnquality goods I believe and that is ~3000 lumens.