Can anyone tell me what kinda runtime can I expect out of the emisar d4sv2 with nichia 519a LEDs, boost driver, and vapcell 6200 mah 26650 at approximately 300-800 lumen setting?
Thanks
I’ll attempt this, I want to get better at “Flashlight Math” -found a 519a compiled graph on Reddit saying 605lm at 2A. Say 10% losses in optics/lens, this is 545lm ish.
Driver is, guessing, 90% efficient, so 2A requires 2.2A.
Cell is 6200mah, so 6200/2200= 2.81hours (2 hours 45min approx?)
Edit: based on ToyKeeper’s reply below, I’m way off, it’s more like 4.5-5 hours.
2 Thanks
Yes, this sounds very accurate.
My D4s running at ~800LM with 5500mah battery is capable of doing like 2h15min
1 Thank
Damn, that’s not very impressive. Zebralight sc700:seems to do 4 hours with a 5000 mah battery, which almost doesn’t make sense. I guess zebra light is more efficient.
I’m looking for a compact flashlight that can do 300-1000 for at least 4 preferably 8 hours.
I know that’s a lot to ask for .
It looks like I might just have to get 2x sc700d
Replies are much appreciated 
I’m confused I just watched this guy
Sofirn sp33 did 500 lumens for 6 hours, this seems to be better than emisar and zebra
The main reason I was looking at emisar is cause I can get nichia LEDs on it
There’s a chance I’ve messed up the calculations… There’s also the discussion of where you call the end to a runtime, for me, I like to look at length of time the lumens at a certain level are sustained for, ANSI runtimes are down to 10% of original output.
Zebralight generally recognised as having very efficient drivers- do be aware when comparing these lights to one another, the LED choice will have a big impact on brightness/runtime: cool white, low CRI LEDs are more efficient than wam white high CRI.
Small reductions in output would result in longer runtimes. E.g. that Reddit 519a graph shows 482 lumens at 1.5A. Running same calculation above, this is OTF lumens of 434Lm.
Driver at 90% efficiency uses 1650ma. 6200/1650= 3.75 or 3 hours and 45 min.
There’s a good review of the SC700d here showing the light does 945lm for just shy of 2 hours, which isnt too bad.
If you want good output efficiency you’d be better choosing a LED with greater efficacy than the Nichia 519a. If you’ve decided you’re definitely a lighting nerd, welcome, and you’ll enjoy the 519a, if you can tolerate the output/runtime trade off.
Edit to add: just to sense check my calculations:
XHP70.2 test (a generation old, but I couldn’t find .3 test) and at 4000K.
638lumens (about 575lm out the front) requires 2.71W (note conversion because the voltage is listed as a 6v emitter in that review).
Let’s use 90% driver efficiency again, the cell provides 3W. The 6.2ah, 3.7V cell has 22.9Wh capacity, so 7 hours…
Are 519a’s really that inefficient?!?!
“There’s a good review of the SC700d here”
Thanks, my review. Excuse all the formatting errors in that review. That occurred when BLF changed their forum software. It can’t be fixed without editing the background HTML.
1 Thank
Well between 300 and 1000 there’s a vast diffrence. Firstly - you won’t find 1000LM for 4 hours from a single cell. Secondly - don’t believe everything advertised.
The Nitecore MH12 Pro has a stated runtime of 8hr 30 mins at 300 lumens, which reviewers (I looked at Zeroair.com) have confirmed. As for the higher outputs, they will not hold the stated output, but as stated above, for a single cell light, it is hard to do so.
Not sure if you can quote from another thread, here goes:
So the 519a will do 99lm/w at 2A for 605lm
The XHP70.2 4000K will do 213lm/w at 1A for 1185lm!
(Same wattage as one is 3V and the other 6V led).
So XHP70.2 is almost 2X efficious.
No I’m definitely not stuck on nichia, I had no clue that xhp are that much more efficient
Thanks
Take this all with a pinch of salt as I’m not entirely convinced of the numbers either! I’ve eyeballed stuff off graphs and done napkin math.
If you’ve the funds, I’d be surprised if you weren’t happy with the Zebra light option, I’ve two older models but highly rate them.
That Sofirn is built to a price point, it’s not in same quality league as the ZLs.
It is cheap and bright though, so if that’s all you need…
1 Thank
For what it’s worth, selfbuilt measured an Emisar D4K. He got 4.5 hours at 480 lumens with Nichia 519A, with a 5000 mAh cell.
The D4S should be reasonably similar, but with the bigger 6200 mAh battery it would be more like 5.5 hours. Or if you use a more efficient LED, it would increase the lumens and/or runtime.
And, of course, if you compare 300 lumens to 800 lumens, the 300 lumen mode will probably run about 3X as long as the 800 lumen mode. So if it doesn’t run long enough, just turn it down one step and it should greatly increase the runtime. Like, he measured level 6 at about 2 hours, and level 5 lasted 4.5 hours. Visually, they’re not hugely different… but the runtime is more than double.
It also helps to look at candelas instead of lumens. A big floody LED makes a lot of lumens but the hotspot isn’t very bright because it’s all spread out. Meanwhile, a smaller LED makes fewer lumens and may be technically less efficient, but the hotspot is quite a bit brighter at the same power levels. If you’re illuminating a target instead of an entire room, the candela spec is usually more important than lumens. An Osram W2 emitter will typically get you more light on the target for a longer time, compared to a large emitter like XHP70.
4 Thanks
1 Thank
Me too, so here are the calculations to be sanity checked, done backwards - to estimate the efficacy of the LED based on measured performance of a flashlight.
These are runtimes of the Nitecore EDC33. The runtimes in this and two other reviews match - the outputs in lumens not so much, but they are close enough for back of the envelope estimates. I’m using the Mid mode (yellow line) for this estimate.
The total measured OTF ‘luminous capacity’ (is there such a term?) is the area under the graph which is roughly 1,600 lm×h.
Assuming 10% loses on reflector/lens, that’s nearly 1,800 lm×h at the LED.
The battery has the capacity of 4 Ah or perhaps some 14 Wh at this moderate draw (or a bit less as the battery still has some juice left when the light terminates).
Assuming the driver is 85-90% efficient, the electrical energy that was sent to the LED seems to be on order of 12 Wh.
The efficacy of the central NiteLab UHi20 Max LED should thus be around 150 lm/W for bare LED and some 115 lm/W off battery with driver, OTF. This as I understand it is about typical for small, undomed, lowish CRI, high CCT modern LEDs?
Am I anywhere close to what’s reasonable?
XHP70.2/70.3 is the king of efficiency, especiall the 70 CRI ones. Also as ToyKeeper said, if you want the longest run time with light on a target, get something like Acebeam L35. XHP70.2/3 + tight angle optics + boost driver is the holy grill.
I’m not looking for throw I’m looking for pure flood.
I like the mule version jlhawaii guy sells but the runtime is not good enough.
I love zebra lights and was planning to get the Sc700d with a luminit filter and just carry spare batteries, but I don’t know how to put the filter on the lens without it coming off very easily.
So I started looking at other lights, and I think I’m settled on the Acebeam E75, cause it’s pretty compact, has medium for 4.5 hours and the bezel can be removed so i can put the luminit filter on it.
Mule just means it doesn’t have a reflector. Doesn’t have any effect on runtime. Unless it uses a different driver
That’s a great help for this thread.
One thing I don’t understand, 4.5hours at 480lm with a 5Ah cell
5Ah*3.7v= 18.5Wh.
18.5Wh/4.5h= 4.11W draw.
480lm/4.11W= 116.8lm/W
Unless I’ve screwed up, that seems like better efficiency than I previously calculated… TheFreeman calculated 116lm/W at 1A from a bare LED, so it would appear the D4K has a very efficient driver and also very low losses from the optics.
But Hank doesn’t do mules with boost drivers, so maybe there actually is some impact on runtimes 
1 Thank