A version with 3v XHP50.3 HI and smooth reflectors would be nice too, for those who want a bit more throw at the cost of 1-2 thousand lumens.
1 Thank
I’d be down with that as long as it doesn’t run into the same problem as that imalent with the 50.3 hi’s and burn one or two.
1 Thank
IIRC they’re fine to 8A when run in 6v, that’s equivalent to 16A at 3v, and you have two of them for each 21700.
Even with P45B it should be fine because of the voltage sag at high currents, assuming there’s enough thermal paste. It might not be a good idea to bypass the tailsprings of a Q8+ with them though.
Ya, supposed to be, but just look at the issues with the imalent. You remember that right? Was a whole thing. Every other one was burning an emitter after a few turbos.
But maybe it’s fixed now idk I stopped following the story
IIRC it was due to a bad batch of emitters and not enough thermal paste.
After those first few lights, the max output has increased (to the claimed 120k instead of barely over 100k) and I haven’t heard from any of them burning out.
1 Thank
HS10 is my favourite EDC. Replacing its driver with the SC21 cc driver counterpart would make it even better.
If possible, a more detailed battery status and a smaller -less bright illuminated- button may be further improvements.
1 Thank
Oh cool. Ya I’m all for it then
Good idea, but I’d suggest making a 3x21700 carrier as well. I don’t trust the longterm availability of these huge cells, but 21700 will always be available.
1 Thank
100% this, but if a more efficient driver is something that’s being considered, at least make it buck+FET. That would be fine, then it’s higher efficiency below turbo, but still has the same face-melting turbo everyone loves it for.
One upgrade might be using SFH55, should push a decent bit more lumens than XHP50.2 as the current limit is higher. Would need a spring bypass and some beefier FETs to make good use of them though.
1 Thank
If built well for using 3 batteries in series, it could be more powerful using buck or boost driver. Then battery voltage’s race against rising led VF, wouldn’t be the limiting factor for max output. Amutorch DM80 is a good example of light with buck driver and high output relatively cheap.
But then forget about onboard charging…
Can’t even throw in a BMS unless you have access to each tap-pointbetween cells.
1 Thank
Yeah, BMS would need to be built in the light itself. No idea has anyone done it yet?
In a Q8 Plus style battery holder one could implement the BMS in the top “spacer” pretty easily. So basically like simons flat top top button top adapter, but with 3S configuration and built-in BMS.
But I would rather keep the cells in parallel, and instead put the LEDs in series for a boost converter.
Here’s how to make the SP36 Peo and Q8 better:
LEDs. I like the sst40 a lot. Powerful, efficient, easy to drive, and okay tint if you get a good one in 5000K, but the SFT40-W is better all around. Offer it in 5000K and 3000K cct.
People like the 519A, and I do, but it’s low output. Not great for making a lot of light. That said, maybe explore some of the other high cri options like the ffl351a or the round die LEDs?
Xhp50.3 HI. Offer it in high cri and R70 versions. Pair it with smooth reflector for the Q8.
Drivers: we all want fully regulated drivers (no FET). I like these, but you take an output hit vs a FET for intial brightness, but higher output over time. Wurkkos did great with the TD01C. I don’t think it would be too much trouble to do a 25 A or 30 A buck driver.
Haikelite HK08 has charging and 4s cells and no BMS as far as I can tell. It’s probably not the safest idea though.
Absolutely horrible idea.
These was a light years (10?) back that did similar, forget what it was, had a bizarre sideways-firing multi LED arrangement on an aluminium (or copper) rod and a custom reflector… anyone remember it? Was it from Flight/ITP?
Anyway, I digress, my point is that it’s 2024 now, internal BMS for series charging should be more common. I had a weird experience with parallel cells discharging mysteriously quickly in a BLF LT1, I now don’t really even trust paralleled groups of cells even though the current dogma is “parallel cells are safe”.
1 Thank
If 1 cell has a self drain, it will take them all down with it. Same thing happens in trucks with 2 or more batteries.
1 Thank
Lights with a battery pack, e.g. Acebeam X75, Imalent MS18/32 / SR32, etc.
This 
Take one out at a time until it stops.
Getting bit off topic, but I know X75 uses boost driver. Any idea what kind of drivers Imalents use?
Not sure. Possibly a boost driver, but it depends on the LED configuration in terms of series vs parallel, as both have 4S battery packs.