Mine too! The cells Neal sells are a super tight fit. Really satisfying, but there is that tight spot in the middle. I have to open mine from the middle of the tube to get the batteries out as a result.
Is yours a reverse or forward clicky? Sounds like a forward clicky from your description. If so, I want to trade! Haha
Does yours have anodized threads on the tail? Mineās threads are all unanodized so nothing short of removing the batteries will lock it out.
Should be unanodized.
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.614b2e8dGujoNr&id=671140993338&\_u=21u0hcib6b54
Nah its a forward clikcy but it looks the same as the Convoy model, no anodizing on the tail threads i just undo it till the battery doesnāt make contact with the tail cap.
Well mine is dead, kind of disappointing i only got to use it a few times maybe 3-4 times. I didnāt get a chance to take it apart when it was working so i could test the amperage. I was testing the output and it hit 24300 lumens which seems to be on par with everyone else.
You can see where the driver let the smoke out under the inductor.
I really hate builds like this there is no centering rings and now all the corners of the LEDs are frayed, i had to undo the front because the wires are to short to pull the driver out without de-soldering the wires.
Last but not least the battery tube done this sigh!
The LEDs are the 6v version.
Wait, itās actually using a buck driver?!!
The inductor must be at the saturation point knowing how powerful the light is.
Yes itās a buck driver with a FET channel. The input is 8.4 volts fully charged so it need to be brought down to 6 volts for lower modes, which should be fully regulated. I tried taking the driver out of mine, but the wires are very short and you have to solder them from the front first.
See, Nightwatch has some very good hardware designers.
Anyway, I think I should contact Nightwatch 
These are some interesting developments. Iāve found most to be poorly regulated _except for Medium, which bewildered me. Knowing the driver has a FET channel and a Buck channel explains that somewhat. Iām surprised that mine doesnāt seem to use the Buck channel for Low or āMoonā though.
Iāll be running test this weekend and that should tell more of whatās going on. The driver is 20 mm and thatās pretty small for a high power light (I saw 51 amps on Turbo using Neal special batteries). Thatās about 300 watts!
Wowā¦.magic smoke. I guess itās pushing physics a bit too much.
I have the JKK76 SFN55.2 (different LEDs of coz) and one thing i noticed in both the lower modes and turbo against say an Amutorch DM70 with Cree 70.2s is that the chinese SFN 55.2 seems noticeably less efficient. So to get more lumens, youād end up with even more heat, and we also know that as we overdrive things and push the envelope like this on a relatively light host (i think the JKK76 has nearly twice the mass), we really hit a brick wall with the law of diminishing returns. Not sure about driver efficiencies between the Amutorch and JKK76, but i guess weād just have to assume that they are very similar.
I already saw that on the video on NWās TB page on the sag, some sort of sagā¦.be it heat somewhere, or battery. Comfortably 30k+ on turn on with the Lishen 21700HP but 2-3 seconds on it dropped to 25k lumens.
The saving grace is that this was 319rmb when it was still available till very recently, i think it sold out ard mid or end of Apr. JKK76 is 538rmb now. (428rmb launch price like 1-2 months ago?)
And if i remember correctly, Nighwatch packages 2pcs of LS 21700HP in it as well, so you are good to go if you ordered from TB (countries serviced by TBās official forwarder CaiNiao can do thatā¦ā¦and yes technically able to ship out as long as you can checkout on TB platformā¦.shhhā¦with regards to li-ion cells. You cannot buy listings that list say 21700/18650 by themselves). With using only 1 cell, i found that the Lishen 21700HP outperformed my Samsung 40T in terms of output/sag at very high current draw on the 3-cell JKK76 during testing, check the JKK76 thread on this. Soā¦.319rmb all in, really incredible value.
ps. sucks regarding the cell wrapping killer for some batteries too.
Just some 2 cents from a regular flashaholic end-user. 
To be fair, theyāre domeless LEDs 
Itās also possible there are more internal bottlenecks.
I just tested mine in industrial enviroment aka powerplant. Lots of poorly illuminated places and shadowy corners. This thing brought light to places that havenāt seen it in decades. Also gets hot very fast. I mean really hot. I added spring bypass to mine, because it was so easy to do cleanly and with these currents it canāt be useless.
Like this:
Possible that if the bypass does make a difference, but it might also do this. Higher lumens before step down period, but faster step down.
I was posting that for the JKK76, if it were able to do well over 33k lumens with good cellsā¦.this might happen. So there is a trade off. With below video, i guess itās sorta confirmed with these chinese LEDs.
edit - also for Audril, at least we can play with and change the thermal limits, whether that makes a difference in reality remains to be seen.
Nighwatch/Jinheng/Haikelite/etc stuff cant.
ie yes, say if we can push the FT03S to nearly 10k, but at what cost. The FT03S is just a wee bit lighter than the NW Chaos i think, perhaps for the head portion where it matters, there is no difference.
Yes, probably faster stepdown. But I think this is just the light for trying to maximize the oomph if only for a short while. There is always dimmer modes if I need sustained illumination.
Do you have before and after bypass figures? Seems like the original spring design is already ābypassedā by that tab, which of course you bypassed in parallel again with that wire of yours. :money_mouth_face:
Or perhaps have you measured on the meter how long in turbo can it hold? Be it before bypass or after bypass.
On this topic, realised now that this light is only 336g.
That means it is in the 8k lumens per 100g range.
Plus the efficiency is somewhat lower than the Cree LEDs.
I measured resistance of the tail spring assembly before and after the bypass. It reduced the resistance for about 40%. I havenāt taken any other measurements.
Wowā¦ā¦.thatās rather significant. Coz i have also done some bypass on my JKK76 but i didnāt do any readings on the luxmeter or be bothered with anything as i wasnāt expecting any resultsā¦.the kind of ājust do it and be done with itā thingy.
The JKK76 steps down in around 30 seconds. Just about at my personal limit of turbo usage, i really cannot imagine anything less than that. Coz after turbo, itās pretty much saturated, canāt do much with the flashlight for a good few minutes other than using it in the lowest mode. At least for the Amutorch DM70, that has a handleā¦if we are talking about using a hot flashlight comfortably for minutes.
What i do if i am at home is to immerse 3/4 of the light into a mug of water if i wanted more play time. :sunglasses: :person_facepalming:
Those resistances are already small, so it might be that my budget meter isnāt excatly the most accurate. Itās Aneng 8008 or something.
Yea, perhaps so. Mine isnāt also, UNI-T lowest range model. So best is to use a luxmeter, that one is very sensitive to differences for what we are using it forā¦ā¦more than sufficient actually. Coz the luxmeter also takes into difference the battery cell sag, internal impedance and things.
Maybe one day i shall remove my alu foil bypass and test it out for such styles of lights. Of coz, not expecting very significant differences to the layman, but like in that youtube vid (Cheuleās flashlights review) it still would interest the crazy flashaholics like us. lol
I know other lights (eg BLF Q8) it does make a difference, just not sure for this style and power range / quality of materials etc