Yea bro, it’s a shame regarding this efficiency aspect. The chinese LEDs are pretty decent in the sheer max output and luminous flux depts, the 2 very important aspects for flashaholics.
I’d not be that concerned if the difference in real life implementations in the flashlight is very small, but from what i found out, it’s not small, though to be fair it’s not insanely huge as well as in not pre 2005 tech of coz. lol! :partying_face:
So just to recheck things in case i made a mistake earlier, I ran the JKK76 and Amutorch DM70 again in 2k lumens levels again last night and verified that both are at the same output levels on the luxmeter (the DM70 is fixed steps and i changed the the JKK76 into ramping mode so that both are at same lvls), 15 mins run duration this time round, the temperature difference between the 2 is rather significant and is not small.
Too bad i gave my IR thermo to my mother in law, if not i’d have temperature figures for a more exact comparison.
Anyway i was reading a Taobao review and found the data below shared by him. Is this data legit and accurate, beats me, but seems to be believable.
Another fella in the reviews section of the Fujian Changda electronics TB listing selling the SFN/SFS/SFH/SFP LEDs also said the efficiency is also lower than Cree.
Oh one more thing, i noticed that the table of results posted above for the 国产70 and 国产50 LEDs are not 3V versions. So there must be some more major thing that is in the equation that is accounting for this.
But I guess to be fair, Cree /Luminus has tons more expertise in all areas, in areas like electrical/manufacturing/phosphor etc.
Anyway, as in the Fujian Changda tb link where the chinese fella posted his review/test, they do have 3, 6,9, 12, 27v etc parts for all the different emitters. (of coz for my JKK76 it’s a 3V SFN 55.2)
heh heh……if you have the dough, why not. Besides, frankly, sometimes i get a bit scared of all these pocket rockets. Small driver, low mass host for the sheer lumens it is putting out.
Just a few years ago some of us were thinking that things like Emisar is too much of a hotrod. 4-5k lumens in 100g+ host.
This is like double (8k+ per 100g), and we haven’t even accounted for the fact that the input power is even high due to the lower luminous efficiency.
After playing with the JKK76 at “only” 26-27k lumens and 490g empty, i think it’s enough. Originally i thought with “direct drive” and 30T or Lishen 21700HP 3000mAh, close to 40k was possible. lol!
Nah….i don’t think that is practical at all, not even for short duration 10s turbo burst. It’d be like LEP usage, nearly useless. But of coz there would always be a rare few flashaholics who can appreciate such output/size relationship.
Yep, according to the vid in the tb nightwatch lisuting, initial 1st second 9k lumens/100g of metal, no joke, after which it slides…8k, 7k. Look at the above table, sbt90 153 watts gives 54 lumens/watt, chinese 70 led is lower efficiency 48 lumens/watt at just 126 watts which is a lower power. Power / 100g is gonna be worse than a cree/luminus light.
Something else gotta compensate for that or else just take the big hit face on.
Thanks for the review.
On TB it is advertised as 30k lumens, which is ok since their integrating sphere + meter does show nearly 30k at turn on and dropped to just below 25k @ 3 seconds.
So i guess we need to bear this in mind when comparing numbers between the few camps. Just take a large part of it as the OTF lumens vs emitter lumens thingy as well. Maybe 15%? hit?
But in real life usages, that is already more than fine enough, already more than sufficient output for the mass of the host. Law of diminishing returns, even for few seconds hotrod usages.
Quote :
How does it compare to the official specs? 22,800lm is lower than the claimed 28,000lm, but not so much lower that you’ll actually notice. That 28,000lm number could be “theoretical” lumens (the output before losses from the optic/reflector & lens). It’s still very bright. Candela of 98,500cd is similarly falling short of the claimed 130,000cd. With that said, it’s still blistering performance.
Whoops second report of magic smoke.
If i am not terribly wrong from what i read on TB, it is directly driven on Turbo.
That’s the “fun” with these budget lights, you’d never know what you are getting after 1-10 weeks of usage. To be far, more branded lights also suffer the same fate sometimes, maybe a little less common.
My first DD light was the DRY triple xm-l driven with 3 18650s around 1 decade ago. Can’t remember how i measured it but i deduced each emitter/cell was doing around 4+ amps range, with good unprotected cells of course (the PCB itself has resistance). During that time there were very few high amps/high drain cells, if any. And the few of us here then in BLF was afraid of the same magic smoke. But still, compared to today’s specs, that’s nothing, be it host mass, heat or the electronics. lol
My advice to other users…… take it easy on the turbo mode. Don’t even rely on the 30 seconds turbo timer. And don’t spam turbo, not even 10 seconds, when the light feels like > 45 deg. Not for the LEDs, but more of for the electronics.
Also, i have this feeling that even at 2000-4000 lumens average usage and the light is 40-50 deg range and heat soaked, do not even attempt to go into Turbo. This feeling is from my experience from other budgetlight previously.
Run it under some dripping water for 15 seconds, it’d help heaps. If you are in winter conditions then that’s great, but it’s the tropics here.
I believe the rule of thumb is a 20% loss from emitter to OTF due to the reflector or optic and the lens. It’s good to keep in mind, particularly buying from makers who are less likely to do actual lumen measurements. The big brands, like Olight or Streamlight for example, will usually do actual OTF lumen measurements. Smaller makers, like Convoy for example, will just do a calculation based on LED testing data of how much brightness the LED should be producing at the given drive current, then they will list that as the lumen specification. There is no substitute for real OTF lumen measurements but theoretical output is better than nothing.
Yeah i have the same “issue” with the JKK76 from Jinheng. I put quotes coz i already know those numbers need to be verified as so it’s not an issue esp for the price, it is what it is, and there is a good chance that i gotta deduct as much as 30%, so do not have any high expectations before getting the light.
In the end i measured, at best, 26-27k OTF lumens range, and maybe 24-25k OTF at worst. It’s not the cells also as i tried with 40T, Lishen 2170HP and 2170LA…for eg 1 pc 2170HP gives me 1400 raw value on the meter while 40T gives 1200 and 3pcs 2170HP gives me 2500 and so does 40T and 2170LA which gave 2500), it’s the driver (it’s not DD on turbo).
But of coz then again, it’s just a ceiling bounce (not even lumen tube) vs quite well known/measured lights like BLF Q8/Sofirn Q8, so i cannot say that my measurement is anywhere accurate as well.
But that is fine, coz it is indeed running very hot in my book already, despite the 50% higher mass than this NW Chaos, the JKK76 is 500g empty, 700g with cells. And my expectation is already just 1 single less than 30s turbo burst and that’s it….not high at all. The emitter efficiencies is in the same ballpark basically.
I was originally expecting 35k to 40k OTF range, coz i thought the JKK76 was DD. But i don’t think that is feasible at all, too much heat (due to the efficiencies involved, this aint Cree) even if the cells can support 1 or 2 bursts.