How do I work out lumens output on AA?

As the title says really. We have the LED out charts and Match kindly did all of them wonderful output graphs and charts. But how do I convert the current to lumens with an AA battery.

So for instance I just received my Sipik SK68 and on an AA battery it pulls 2A, so how do I convert that into Lumens using the charts?

Thanks,

Marc.

The easiest way would be to measure current at the LED, but going off of the info you have told me I can make a rough estimate.

2 amps at 1.3 volts is 2.6 watts drawn from the battery. 2.6 watts times 90% to account for efficiency is 2.34 watts. So there our rough LED wattage is. To convert that to LED current you would divide by 3.2 so .73125 amps are going to the LED.

I new you would answer me Scaru! :bigsmile:

So my little Sipik is pretty damn bright then! I worked it out by dividing the amps at the tail cap into 3 (3AA being 1 18650) so I wasn’t too far out. But I wasn’t sure if I was right doing it that way.

Thanks Scaru,

Marc.

Scaru, he da man, da man wi’ da Mycroft touch, he know so much.

Hope you don’t mind Scaru. I’m feeling punchy this morning and that Olight thread has unleashed a monster.

But that leave out things like the optic efficiency (light loss in reflectors and lenses). This is typically around 30. And much more in zoomie lights. And then going from flood to zoom takes another 50 hit.

Also cheap AA boost drivers are seldom 90% efficient. You are lucky to get 80. I’ve seen them as low as 50 (and one was around 14%!!!)

To be honest I was only looking at emitter lumens. But your right, especially with the Sipik Sk68 I just received. It must loose half it’s lumens when zooming. But it is bloody good on flood, in fact I would say as an estimate it must be give close 140 lumens out the front. It’s bloody bright! That’s on AA as well.

Marc.

The formula for working out the LED current has four variables of which you supplied one. It's impossible to know the answer without also knowing the remaining three variables. Here's the formula:

LED current = (battery current x battery voltage x efficiency)/LED voltage

In addition to the battery current, you need to know the actual battery voltage under load, the actual LED voltage (not the typical figure supplied on the datasheet as the actual figure can vary quite a bit), and you also need to know the efficiency of everything in the path between the battery and the LED - this includes the switch assembly, the host, the pill, the driver, and wires.

Determining the lumens from the LED current isn't simple either. Even if you just wanted to know the emitter lumens, you would need to know the LED current and also whether the LED is suffering heat sag. Tests like the ones done by Match were performed with the LED on massive heatsinks ensuring optimum heat dissipation. The heat dissipating properties of a SK68 cannot compare.

Figuring out output that way is pretty hard. But a good rule of thumb for most AA powered flashlights is that the average 1xAA driver supplies about 400mA to the LED and the average 2xAA driver supplies about 700mA to the LED. Some do better, but not by much, and the drivers in most SK68 don't really fall into the "better" category.

So, for your average Q5 SK68 or clone, you're probably looking at something around 120-130 emitter lumens.

Using tailcap current to estimate output is something that's only really useful for lights that use linear regulators and a single Li-ion. The formula for determining LED current for that setup is a lot simper. It is:

LED current = battery current x efficiency

Voltage gets eliminated from the formula and really simplifies things.

Thank you for that, very comprehensive. The only thing I would say is that there is a lot more than 120 lumens at the emitter, OTF yes comparing to my various other torches, which are ANSI rated. I used to have a LED lenser and on flood it is almost the same, maybe slightly less but barely noticeable.

I may have got lucky of course, but I hear people on CPF saying these things give 60/80lms, well not this one that’s for sure!

Marc.

Have you compared to an ANSI rated light other than LED Lenser? LED Lenser's ANSI figures are usually exaggerated, sometimes wildly. They are among a number of companies, ZebraLight to name another, who's ANSI ratings don't measure up when independently tested.

ANSI oversee testing standards but the organization doesn't oversee the actual testing itself. Companies usually do that on their own, so companies can cheat and exaggerate their numbers and still put the ANSI name on it if they want as long as they "followed" the standard procedure for testing.

Fenix's ANSI numbers are pretty legit. Can't think of any others off the top of my head, but there are a few. There are even some companies that under report their lumens a little bit.

+1 to all that.

If you have a DSLR, you can use the light meter in it to measure lumen output of any flashlight, any mode. I find eyeball comparisons so deceiving once different tints and beam patterns are involved - ceiling bounce or not. I measured a “220 lumen” NEBO Redline at ~40 lumens flood, ~25 lms throw, and had no idea it would be that far off.

In my collection of sub-lumen/1xAAs Foursevens, Eagletac and ThruNite all seem be working off the same lumen scale, while Zebralight is off (well, at least from the other three Co’s) by 30-50%.

I don’t think the Lenser is ANSI rated, It is however rated at 200 on turbo and 175 on high, I was just saying that there was not much difference with the high mode. But my Klarus is ANSI rated and whilst the Klarus IS brighter, I can assure you that the Klarus is not double the difference. I am not saying my SK68 is 140Lms, but I am saying it is definitely not 60.

@Reppans, I did not know you could do that with a DSLR. Great info!

Thanks,

Marc.

Refer to Rojo’s points about “ANSI rated,” from some of the tests I’ve seen, Klarus uses an ANSI scale similar to Zebralight. :slight_smile:

Not sure if you’re aware of the logarithmic human perception thing, but in my experience with a light meter, if it looks anything like “double the difference” you are getting into 4x more lumens. IMHO, especially when you account for different tints on a ceiling bounce test, if you can perceive a brightness increase at all, you’re in the 50%+ more lumens category.

Never heard of any problems with Jetbeam/Klarus when it comes to ANSI, they are always spot on from what I have read here and on CPF.

The fact remains that it has around 180LM at the emitter and that is good enough for me.

Marc.

Yeah…I guess it’s purely a perspective thing. I tend to grade on a “curve” while Selfbuilt seems to work more on a “lowest common denominator/no child left behind” basis. But politically, I understand why he has to do that. I guess, it’s just that most of my lights in the AA/sub-lumen class (from which I benchmark) are so significantly considered “underrated” by Selfbuilt… by around 50%.

BTW, his numbers are not official ANSI - he uses a “best fit line” which is discussed on his website. To me though, “best fit” implies a roughly equal number of overstated and understated lights and, at least in the 3V and below class (AA/AA2/123) with which I am familiar, “understated” lights are very common, while “overstated” lights are really quite rare in his reviews. Using a “curve,” I guess you could say I would “fail” a lot lights that he “passes.”

Anyways, the important point is that Selfbuilt’s numbers are at least accurate on a RELATIVE basis, meaning that if you wanted to compare a Quark AAX, D25A Clicky, or Neutron 1A apples-to-apples to an SC52 or ST10, you could use his understatement ratios to bump up the specs on the former three by around 50%?