4Sevens Design Philosophy #8 - Always OUT THE FRONT Ratings

[4sevens]

There are lots of ways manufacturers of lights specify the performance of their lights these days. Most take the borderline-dishonest route and of inflating their numbers. We've decided from the very beginning that we will take the high way and specify conservative numbers from measured OTF (Out the Front) Calibrated IS (integrated sphere) readings. Our goal is to under-promise and over-deliver. It's not worth losing customers by deceiving them about the output. We're here for the long haul.

Technically there are at least three ways to specify the output performance of your product.

1) Measured OTF IS readings. Also, known as "torch lumens." It is the amount of lumens expect to leave the business end of a flashlight.

2) So called "blub lumens." Lumens that come from the LED. After the light hits the reflector and lens or goes through an optic, there are unavoidable losses. However, citing lumens at the LED gives manufacturers a way to state higher lumens that what is actually coming out of the flashlight, giving them an edge over other brands - at least to those who have not been informed. Whether it has been measured at the blub has also yet to be determined.

3) Then there are outputs based one drive current interpolated from LED manufacturer's specifications. This ones the nastiest of them all. Flashlight makers will quote an output based on the PEAK drive current of their circuit matched with the highest possible lumen rating for that specific LED. LED's are binned within a range, yet these manufacturers will claim the highest POSSIBLE output. Also, they don't account for the fact that the LED manufacturer's spec is at certain junction temperature with a certain ambient temp. They just take the best of the best case. Usually this gives the MOST inflated lumen rating - even more than "bulb lumens"

I won't include those who just flat out LIE about their lumens. Just take a look at ebay.

Anyway, I'll stop ranting.

Lumens Out the front FTW!!!

[jabe1]

excellent.
waiting...

[Zeruel]

Good good.... now give us the babies already.

[AardvarkSagus]

This is one of my favorite design philosophies thus far. Great news for us customers!

[Haz]

That's good!, i support this too

[adirondackdestroyer]

7777's,

Quit torturing us and show us these bad boys already!

[NonSenCe]

HooRay!

[csshih]

indeed.

this "otf" lumen testing should be heavily advertised, though, or you might lose some uneducated customers..

many fall prey to the 300lm ebay lights.

[Cosmo7809]

You mean my china superfire is not 1000 lumens?!?!?!?! WTF

[Toohotruk]

[cave dave]

Originally Posted by 4sevens

2) So called "blub lumens."

Blub Lumens?

Besides the spelling errors in a couple places, bulb lumens are just a subset of (3) poorly interpolated lumens. No flashlight manufacturer is sticking a bare LED into an IS and measuring the results.

Another method that not many seem to be used is:
4) Mathematical estimated lumens, this would take the spreadsheet values and calculate the output including junction temperature losses (data available on cree datasheets) and reflector and lens losses, all of that data should be available to the manufacturer. If the manufacture doesn't know the loss his reflector coating is producing I would worry about the quality of their lights.

Will you be purchasing your own integrating sphere or renting time like Arc does.

Also how will you account for the LED lottery variability? Avg or min? How many units will be measured to establish a baseline?

[polkiuj]

price?
=P

[Choaticwhisper]

Originally Posted by polkiuj

price?
=P

Stated in another thread, under $100.

[Loanshark]

Okay okay, when??

[Choaticwhisper]

I think this is the best thread to post my question in.

I seen in the one of the photos that the 1xAA light can handle up to 4.2v so using a 14500, Do you get 90lm and less with a nihm?
Or nihm get 90lm and 14500 get more? More or less runtime?
Nothing changed?

[Zeruel]

I wonder what will OTF 90 lumens be in "fake" figure.

[Choaticwhisper]

Well +20% is 108. So either 110 lumens or 130 lumens, Just to beef it up little bit.

[Zeruel]

Originally Posted by Choaticwhisper

Well +20% is 108. So either 110 lumens or 130 lumens, Just to beef it up little bit.

Not too bad.
But I was hoping for Surefire's kind of estimation

[Choaticwhisper]

Originally Posted by Zeruel

Not too bad.
But I was hoping for Surefire's kind of estimation

Well Acording to MrGMan's IS thread surfire was just about dead on with the p60L and the the E1B, I dont remember the others.

[Zeruel]

Originally Posted by Choaticwhisper

Well Acording to MrGMan's IS thread surfire was just about dead on with the p60L and the the E1B, I dont remember the others.

Ah... perhaps I should specify E2DL.

[4sevens]

Originally Posted by cave dave

Also how will you account for the LED lottery variability? Avg or min? How many units will be measured to establish a baseline?

We've quoted these ratings as "typical" and "actual measured." What we've done is take a dozen, measure them in an IS and take the lowest output. We're not checking whether a specific LED is at the upper bound or lower bound for the R2 bin. Thats way to tedious - plus there are other variations such as ambient temp, humidity, how many hours on the LED etc etc.
We just measured a bunch and noted the lowest number.
Runtimes with 123's were done with name brand cells. For AA models we used 2500mah nimh.

Originally Posted by Choaticwhisper

I think this is the best thread to post my question in.

I seen in the one of the photos that the 1xAA light can handle up to 4.2v so using a 14500, Do you get 90lm and less with a nihm?
Or nihm get 90lm and 14500 get more? More or less runtime?
Nothing changed?

Excellent question. The 90lm rating is based on a 1.5v input. If you do put a 14500 cell in there (which has 4.2v), it will put out a regulated 170 lumens!!
The circuitry is the same for the AA, AA-2 and the 123. They
all have an input voltage of 0.9-4.2v and the output is regulated by a buck/boost circuit.

[Choaticwhisper]

Originally Posted by 4sevens

Excellent question. The 90lm rating is based on a 1.5v input. If you do put a 14500 cell in there (which has 4.2v), it will put out a regulated 170 lumens!!
The circuitry is the same for the AA, AA-2 and the 123. They
all have an input voltage of 0.9-4.2v and the output is regulated by a buck/boost circuit.

Well thats just great, Now I will have to give you more money. I was setting on ordering the 2xAA model, I just might have to have that 1xAA also.
But reguardless Im waiting on the warmer tints, Maybe.

[Justin Case]

Originally Posted by 4sevens

We've quoted these ratings as "typical" and "actual measured." What we've done is take a dozen, measure them in an IS and take the lowest output. We're not checking whether a specific LED is at the upper bound or lower bound for the R2 bin. Thats way to tedious - plus there are other variations such as ambient temp, humidity, how many hours on the LED etc etc.
We just measured a bunch and noted the lowest number.

But as you note in your very first post, output depends on junction temperature (and ambient temp). Especially in small lights, the heat sinking is most likely not going to keep up with the heat generation, and the junction is going to heat up by some amount. Thus, do you measure OTF lumens on initial turn-on, OTF lumens after X minutes of continuous run time, OTF after X cycles for a Y% duty cycle, or what? I can build an MC-E lumens monster out of keychain light that can put out a huge OTF lumen number, but it won't be practical. If you measure the lumens after a few seconds of run time, the OTF figure probably will have dropped by 90%, followed shortly after by burnout.

Bottom line is that OTF lumens is also not the be-all, end-all figure of merit.