Firstly I don”t consider myself a modder but I can solder 
My question.
I have a Fenix HP10 headlamp that runs on 4 x AA”s.
The cells are in series so 6v & I believe the lamp uses a 3v buck driver.
The Duracells that were in the battery box have leaked & destroyed the contacts so I am looking to fit a new battery box.
Now if I used another 4 x AA”s in series battery box I would use lithium primaries as these will not leak.
I have used lithium primaries (Energiser) in other applications & found that the cells voltages are around 1.80v & have not caused any problems.
If I use 4 x AA lithium primaries in series that will be about 7.2v & I don”t expect that to cause any problems.
However I am wondering what would happen if I used a 2 x 18650”s in series battery box ?
This would give a nominal 7.4v or peak 8.4v so a little above the 4 x AA voltage but I am assuming that the driver would cope with the extra voltage fine ?
I have some 2 x 18650 in series battery boxes handy & plenty of good quality 18650”s but thought I would ask the advice of those more knowledgeable before rigging this up & potentially burning out the driver ?
…or just buy a quality, low cost 18650 headlamp? 
(You can tell I’m not a modder)
I would if I could but see this thread Headlight recommendations wanted
It seems that none of the big brands makes what I am after & I am looking to resurrect my old HP10 as a back up to an Olight H35.
It’s hard to say if it will work or if it will work but burn out after a few uses. Do you have pics of the parts? I bought a headlamp that is this design but uses 2 18650s from fasttech look similar.
The skilhunt H03R with coupon is sbout 36$
Or nitecore HC60
Both really nice 18650 lights
The skilhunt is extreemly light with only 44g
Thanks for the recommendations but I would like to breath new life into my HP10 if possible, preferably by using 2 x 18650”s but if not 4 x lithium AA”s.
If you look at my headlight recommendation thread here Headlight recommendations wanted you will see that if I were to buy another headlight that neither of your suggestions meet either my beam pattern or run time criteria.
I can understand wanting to breathe life into it. But unless it’s some newer model I believe it has a Q5 led which isn’t also that great but if you like the profile of the beam and everything you might as well just keep it simple and find a new power bank without spending any more money.
How does it work with a single 18650? I’d imagine if it does indeed work the brightness may not be all that different so long as it’s holding at around 4v, in which case a 26650 may provide adequate run times. Also, if it does indeed run from a single 18650, it may not be a good idea to use two series cells simply because if left on the light could drain the cells down to an unhealthy level.
You may well be right on the emitter being a Q5, it certainly isn”t an XML.
I can convert it to run on 18650”s, if it indeed will run on them, for nothing as I have battery boxes & plenty of cells.
If I were to convert it to 18650”s I would probably use protected Panasonic 3,400mah ncrb”s as the head lamp does not have LV protection or warning so over dis-charging would not be an issue.
Although not very powerful in terms of out put compared to today”s headlamps, with the combination of a throwy beam profile & flip down diffuser it is very usable & seems a shame just to discard it.
Also it will only be kept as a back up to my main head lamp.
ah good point, protected cells should be fine. as for the driver handling the heat of 2 series cells, not sure. definitely seems like a tricky risk! I wasn’t able to successfully find any details about the driver board other than the IC specs. it does supposedly have overheating protection but I can’t determine if that’s just in the form of a timed step-down from high or some kind of thermal regulation in the driver. either way as buck drivers go I would imagine there would be some leeway with the input voltage. not like it’s some actual solution, but maybe you could try it with a couple CR123’s giving you a little less than 7v total and just see how it works out heat-wise before trying series 18650’s.
It has already run fine in the past on 4 x Energiser Ultimate lithiums & they are 7.2v so I know that it will handle 7v fine.
I am just a little concerned whether it would handle the potential 8.4v from a pair of fully charged 18650”s.
Only one way to find out if there will be any “magic” smoke I suppose & that is to try it.
My instinct is to test the light with a single 18650 to see if it performs well then modify the 2X 18650 holder for the cells to work in parallel. The possible downside of a single LiIon is if the light has low battery warnings or protection which should activate around 1.0V per cell, or 4.0V overall.
ONLY IF YOU ARE WILLING TO TAKE THE RISK of destroying the light, connect two series LiIon cells and quickly put on then take off the light to see if output is normal or significantly brighter than normal. If significantly brighter, obviously, 2S, LiIon is too much voltage for the driver.
I suppose the best thing to determine is how hot the driver gets from the 7.2v or so using the lithium AA’s. I’m still not quite sure what kind of driver it is. If it’s some cheap linear regulated driver then I’d imagine it would get quite hot from 7v, and 8v would seem to be pushing it way too much and burning off too much current. If not, and like you say it seemed to be doing fine at around 7v then it may be a pretty good buck driver that can handle that 8v without much issue.
Fenix are a reasonable quality light so I would assume they have used a reasonable quality driver.
The headlamp does not have low voltage warning or protection AFAIK.
The only thing that it does have is a timed step down on the turbo mode.
If the light is designed to be fed 6v, & performs fine with 7.2v, surely using 4.2v, assuming a fully charged 18650, won”t be enough will it ?
I don”t know much about drivers as you can probably tell but isn”t the use of a buck driver to reduce the input voltage to the led ?
That way you can run run several batteries in series that have a voltage above what the led is designed to run at but still deliver the correct voltage the led ?
Then as the cells deplete & battery voltage drops the buck driver limits the current less & less until at about 3.0v input voltage you basically have direct drive ?
4x 1.5V or 1.8V is already above what the LED needs so this has to be a buck driver and 4.2V is also above what the LED needs, so voltage needs to be reduced for it to power the LED safely. If you have any 14500 or 16340 cells, test with one of those so you wouldn’t have to wait hours to see approximate runtime. I’ve played with the drivers of fenix HL20 and HL21 headlamps already and being designed for a single AA cell, I presumed that they would be fine using 2AA cells. WRONG. They both went into high mode only and did nothing else. The point being that fenix designed the driver with a specific input voltage range and I’m being cautious assuming that 8.4V may be too high. Again, it’s only an assumption till you perform a test.