Climb up on the podium, go ahead, collect the BLF Gold Medal. 
Nicely done!
Yeah, I guess you are right, or at least partly right. It says âworking voltage: 6 - 12Vâ, but it also says âOutput power: 3-4 li-ion battery LED current 2.8-3A, 70-100Wâ.
I have not used it much with 4 cells, other than breafly testing it, but I did retest it last night, and I measured 3.5A LED current and 7.2A tailcap current from four KKs 90% charged. Cells must have been low on voltage last time I measured. Indoors it became uncomfortable hot after just 60 sec runtime, but outside in winter temp I ran it 5 min before I quit. High temp protection did not kick in either way.
Practical, it is not for sure, but maybe a heavier host like the X 100 would help. At least you got your 8000+ lumens measurment :bigsmile:
With a little more realistic power source, I used 4 KeepPower 26650's (protected), freshly charged at 4.19-4.20v, measured 7.04A tail:
lumens: 9,078 @start, 8,636 @30 secs, throw: 87.5 kcd
Focus is still not perfect - still have a ring around the center hot spot. Think I know what happened of why the 2nd J18 had more reflector/LED problems than the first. I used SinkPAD's on the first, Noctigons on the 2nd.
Distances between solder pads on 16mm SinkPAD: 11.25 mm
Distances between solder pads on 16mm Noctigon: 10.0 mm
This little difference has made a big difference in getting the plastic alignment pieces seated good.
If you use insulating tape on the reflector bottom and/or on the solder blobs, you can often skip them alltogether or possibly use just one or thee of them. Some reflectors, like the shocker and other flat bottoms, you must machine room for the solder blobs, and then insulate.
Umm, I think I invented the Shocker machining out space for the wires here: https://budgetlightforum.com/t/-/19641 
. Yes, I know various technniques, it's just very time consuming to re-work 7 LED's, and they still need to be properly centered in the holes the LED doesn't fit tightly. Also, all 7 are sitting on AS5, so they are floating somewhat, easily can move, etc. - that's good and bad. Good they will adapt and re-position easily, bad that they move easily . I'll have to get back to it at some point, but got several other things I'm working on at the moment.
Uh, yeah. Tom kinda knows what heâs doing with these multi emitter lights. Tom has modded so many BTUâs and spent so much time perfecting them, he likely sees them in his sleep
GBD - "invented" is a strong word, but think I was the first to come up with it.
Anyhow, thanks for the tips - I know it was very well intended! I did look at it from several angles, just really should have spent more time on it.
I prefer Fujic in multiemitters, dont think it matters much if done properly, and it makes life easier.
I did drill and tap holes for two fixing screws to each emitter on a bighead 6x just to be able to use a super noncuring compound, bot I wont do it on 12x, and most likely never again :weary:
Maybe even the wheel was invented more than one time? 
I've been staying with AS5 on the multi emitters. The wiring is a little hairy as they move around, but if you can build in some slack, it's not so bad. They are normally held down well by the screwed down reflector - much better when you have 2 screws vs. 1.
I have some difficulties working with 3 LEDs mod (i.e. Shocker) and I canât imagine working with J18! Think I will probably mess up all that LED positioning and wiring.
Sometimes I mix AS5 with Fujik together to solve the moving around problem. I apply AS5 in the middle of the star and a little of Fujik on the side. That way I can freely position the LEDs when settling in the reflector, and when itâs done I would just wait for the Fujik to dry and hold them in place.
May I know is there any problem with the KD 9 LED driver recently?
Not for me, mine are sitting uneventfully in the electronics junk pile currently, after failing. Third one was not a charm.
Hi B42, how does the driver fail? It doesnât function at all or it doesnât deliver the expected current to the emitters?
B42 can let you know how his failed, but I've experienced two failure modes on the KD drivers, both 7x and 9x, I've seen fail: the first mode is complete failure, the second mode is losing the ability to switch modes. The driver that lost the ability to switch modes was stuck in the lowest mode, which is probably what kept it from frying itself.
They all put out a lot of amps, and none of them lasted more than a few minutes, even with cautious use.
The failures I've seen (2-3) have been complete failures, but many more failures I've heard of posted. Somehow my 2 modded J18's are still surviving - one with the 7X driver using 6 XM-L2's on copper, and the other with the 9X driver using 7 XM-L2's on copper. They get maybe 30-45 secs runtimes at most, not always with the most high performing cells.
Once the KD drivers fail, I at least have the option to replace with a zener modded driver and wire up the LED's for 3 cell operation, using 6 LED's total (2P3S).
You can also get away with running all 7 emitters with a zener modded FET driver. I just paralleled the center emitter with one of the strings on the J18; I also did this with the 5 emitter J12. I had to try it and both lights have been running that way for months now with no problems.
Dang!! That sounds the best way to go - makes sense to parallel in the 7th. Wish I had more time to play with these.
Recently put in a zener modded BLF17DD in a 9X TrustFire light (same as ebay.com/itm/TrustFire-9-x-CREE-XM-L-T6-LED-5-Mode). I'm getting crazy amps out of it with so-so results, limited because of blocking LED alignment pieces and XML's on stock stars. I disabled voltage monitoring though til I can figure that stuff out. Think I got 11A at the tail, so about 3.5A per LED. I like this light as a host though - massive head, seems to be good heat sinking but the stupid segmented pieces for the battery tubes don't line up well, so protected cells are a PIA to drop in to run 3 cells. I paid only $45 each for 2 of these 9X lights at the time - great deal, better than J18's for sure.
I still like the J18s more than a lot of the big head lights. They are light, well balanced, and transfer heat well for their size.
I have built a few hot smaller 12x and 15x lights and they just can't handle the heat well. One of them was like the light you linked to, but with a 12x emitter head. They almost all have massive heat sag (with the stock aluminum MCPCBs). They are good 5000-6000 lumen lights, but I have to limit the output on them.
Forget #'s I got (maybe 5.5-6.5K range), but yes - this 9X light has a huge heat sag, plus battery sag I assume at an 11A draw initially... Think it dropped bout 1,000 lumens in the first 30 secs.
Interesting bout the J18 results you got. I agree - very light weight but the head sure heats up fast, but that's probably a good thing .
My nephew really wants one of these type of lights - he's a NYC Port Authority officer, but wants it more for his place up on Vermont. I'm sure he'd be doing some showing off as well... I'm sure I can instruct him how to use it safely, and set him up with good cells and a good charger. For 3 26650's though, it is a PIA for charging them. The i4's I believe fit only 2 at a time (that's how I always did it), but I have a ton of chargers for my own use.
What's the details of your J18 builds? Do you have LVP working? If so, what R1/R2 resistors, and what value do you use in the firmware? I know you can do the math, but I'm still unsure how this all works. I've fiddled with setting up my own testing of LVP thresholds but don't trust the results.
I'm thinking hottest setup today is XM-L2 U3's SinkPAD/Noctigons, Zener mod FET based driver setup for 3 cells. Limiting amps with lots of 7135's may be a good idea for the one to my nephew, but not sure of resulting amps/lumens then. Not sure what to expect, but a 10A tail reading for example, 120 watts of power sounds pretty insane -- over 4A per LED. You can control a FET based driver somewhat by the cells you choose as well.