which flashlight brands have the most efficient drivers?

You can have a 6000lb SUV with a V8 thats efficient.

Most high MPG cars are small. If I have to give up size to gain in MPG , that does not make it efficient. I paid for greater mpg by giving up size.

With electronics, efficency has to do with how much energy is lost to heat. So if you have a 1000lm light that gives 1hr run time compared to a 1000lm light that gives 45 min runtime but runs much hotter, that light would be less efficient.

If you compare a 1000lm light that runs for an hour vs a 2000lm light that runs for 2 hrs, you really dont kbow which one if more or less efficient.

if I have a 2000 lumen light and only use 1000 lumens, that will double my runtime efficiency :wink:

an AA light cant reach 1000 lumens at all, it has zero efficiency, for that lumen level.

similarly with cars, MPG efficiency changes depending on how fast you need to go, or how heavy a load you need to move

jon me and most humans dont speak the same langauge lol. i want as much lumens per watt. i use low lumens most the time under 2 lumens.

true
if lumens per watt efficiency is what you want

ignore me and anyone who tries to talk about runtime efficiency, Lumens, and battery size… lol

Im glad you found your Zebra AA suits your needs and wants

enjoy it :slight_smile:

I like topics threads that have high knowledge shared efficiency. Like this one. This is a very high knowledge sharing efficiency thread. But there is room for other types of topic threads, for example, the recurrent topic that reminds me to wash my dirty laundry.

Well, anyway, I have seen the names of many various manufacturers mentioned for good, efficient drivers. I certainly don't know of any makers that are known for horrible drivers.

Efficiency has a ton of variables. A driver might be efficient at high outputs but not low - or vice versa. Also, you can have efficient drivers with inefficient LEDs, or the other way around, etc…

A FET driver at full blast is the most efficient driver there is! Too bad the led is far from its efficient operating zone, though.

1 Thank

A post was listed recently asking why Candlepowerforums is not as active as it used to be. I have to say it’s because of rude responses like the last three on this post.

Technology like any electronics is changing and sometimes improving year to year. For example the driver from the Emisar DT8 is just now starting to be used in Emisar D4V2 and Noctigon models. There are tons of variables yes, but that’s a more efficient driver than what was previously used on both D4V2 and KR4 models from any previous year.

Not only will new folks always be moving into a hobby, but technology advances , so “use the search” or “this always gets brought up” is not very welcoming. List some personal advantages you’ve experienced or just save your energy for something positive. If you think you’re done learning, you might just be the fool yourself.

1 Thank

Output efficiency depend on:

  1. Driver topology and quality;
  2. Led specs- lumen per watt output;
  3. Led current;
  4. Contact - wires in Ω;
  5. Optics clearness;
  6. ……

This info should be available before buying , but sellers advertising only “mooore lumens”, and yes, you want moore lumens, right ? :laughing:

It’s the same driver (the one that was introduced with the KR4), just with different current settings (5 7.5 and 9A), this has a negligible inpact on efficiency and in the end it’s still a linear driver with subpar efficiency.

im weird im not a lumen lover i mean max lumens is not what i look for. my fav feature are small size compared to the battery it uses and lot of runtime for the cell comapred. i guess you call it lumens per watt

The SBT90.2 seems to have relatively low efficiency at high outputs but at 1000lm efficiency appears to be 164lm/W for the emitter itself according to djozz’s thread . If the Fireflies T9R driver is >90% efficient as claimed, then at 1000lm output it should be capable of at least 148lm/W with this emitter. That is of course not considering losses in springs, battery resistance, etc. which may be significant. Using 1lumen’s measurements , it looks like it manages 137lm/W @ 1012lm assuming this was tested with battery at 4.1V.

The combination of throw, large hotspot, decent spill, moderate sustained output, FET turbo capability, high efficiency, balance in the hand, and Anduril 2 means this checks off a lot of boxes both as a practical and enthusiast light. Makes me excited for the T1R if they can deliver something like this in a pocketable form factor.

Maybe I’m in the minority here, but I consider the T9R to be the ugliest light I’ve ever laid eyes on. Even if it were the greatest light ever made I don’t know if I could bring myself to buy it.

You’re probably in the majority but I quite like the look of it, at least in pictures. Hopefully it lives up to expectations when it arrives.

Olight drivers are the most efficient out there but unfortunately they only release CW and don’t care about high CRI. Fireflies that uses the LUME driver are also 90+% efficiency but they had problem sourcing chips for the driver since the pandemic.

First off, thanks for posting some difficult-to-obtain information on the other thread , I’m a ZL fan, but never had any modded, or tried to take them apart myself, so the details of their “legendary efficiency” has always been a magic black box. I’m an EE, so it’s definitely interesting to have some data sheets to read, now. I’m not clear how they do the CC control with those chips or what other cleverness they might have, but it’s cool to know something.

It’s interesting that there’s at least one level that the 53’s spec shows it as more efficient. The 53 says 30lm for 21h, the 5 says 21lm for 21h. I have my doubts about the 30lm.
At the lowest high the 53 spec is 122lm/4.2h=29 lm/h vs the 5 at 166/2.8=59lm/h. Kind of suspect, since if the 53 is ~60-70% efficient, then at best, a 90% efficient driverf would give you 50% more lm/h, not double. Maybe you can’t use the numbers that way for an accurate comparison.

I picked up a second SC5 since it seems to be on the discontinued list, now, though still in stock. I didn’t know the chip was that much niftier, so I’d like a spare.
My 53w will be in the shop for 3-4 months on warranty repair, as it seems to have a problem on M1 where it’s drawing 3x the current (~1.2A vs 0.4A) vs M2 instead of the H53 and SC53c where it’s double (~0.8A vs 0.4A), and rough light-meter measurements tell me the ratio of M1 to M2 is about 2:1 for both lights. M1 and H4 are probably my favorite modes. I’d hate to be burning off 400mA for nothing at M1. Not sure what could be wrong with the circuit to cause an efficiency dip that bad in the middle of the range, but for $4 shipping, I’ll get it fixed.

I think you kind of hit the nail on the head, here. The consumer group of LED flashlights that’s electrical circuit savvy is so very small. Most flashlight makers are interested in profit and their target consumers are not engineers. This is why Emisar/Noctigon is so very much appreciated here, as the man behind the brand, Hank, is active in the community and responsive to those who are electrical+LED savvy.
At this point, it seems to me that LED output capacity is in a golden age—many efficient LED choices now, relative to 10~15 years ago. But of course, the trick is mating them up to a well designed driver circuit and firmware.
Jon Slider makes a great point—it’s not just the circuit design and LED, but also how you drive it. I think 1000 lumens is a good benchmark. THAT would be interesting to see. A kind of shoot-out on efficiency across many brands with single cell driver; how long they can produce a sustained 1000 lumens output before stepping down, in a moderately low ambient temp environment.

Have you measured or know the efficiency of the Dr Jones H17Fx? Specifically with SST-40 and SFT-40, which is my current EDC in a Convoy S16.

It’s a linear (1x7135+7x7135) + FET driver, the linear part is like a variable resistor, it adjusts its resistance so that, at the desired current, it drops the voltage down to the Vf, so the efficiency (%) is Vf/Vin.
For example SFT40 Vf at 3A (full linear current) is 3V, and the cell is (a 30Q) fully charged, at 3A it drops to ~4.1V, plus with the springs+switch to ~4V, that’s Vin, 3/4= 75%.
At 3.6V, drops to 3.4V, 3/3.4= 88%, so the efficiency increases as the cell discharges.
Because there is PWM dimming, there is an LED efficacy (lm/W) cost, at 50% for example the LED is ON half the time at 3A (let’s ignore the 1x7135 channel), so when it’ ON the efficacy is closer to the 3A efficacy than 1.5A efficacy, but not exactly 3A because it doesn’t run as hot, this is difficult to estimate. At 350mA (1x7135) and below there is no efficacy cost because the LED is around maximum efficacy at this current.

For the FET part, the driver efficiency is again Vout/Vin, but because a FET has low resistance, the efficiency is high, however, because of the very high direct drive current, with PWM dimming the LED efficacy cost is large, again this is hard to estimate, the only way to know is to measure the efficacy of the entite light.

3 Thanks

Thank you for that explanation.

if you really want lumens per watt, you have to also look at led efficiency

usually the cold light blue ones have the most

it;s a tradeoff