Establishing a runtime standard

[RonM]

I was wondering if we could all agree to a standard of measurement for runtime? Most manufacturers seem to define it as the point where the flashlight no longer puts out "useable light." Very vague and subjective.

I propose that runtime be defined as:
The point at which the flashlight's output is one-half of the initial output with new batteries.

This would be similar to the way MTBF is determined. MTBF is the point at which half of the devices have failed. I'd also recommend that "initial output" be measured after the flashlight has been on for 60 seconds.

So what do you all think?

 

[MarcV]

Sort of half light instead of half life as is common to describe radioactivity?

 

[Roy]

How is light output to be measured? I use a photo cell and a multimeter and measure the mv output of the photo cell. I place the front lens of the light seven inches from the photocell. Why seven inches you ask?

That's how far the edge of my computer desk is from the wall! I've placed the photo cell on the wall so that the beams from the flash lights center on the photocell when the lights are laid on the desk.

 

[The_LED_Museum]

Using that methodology, some flashlights will have a "battery life" that could be measured in SECONDS!!

 

[remuen]

Originally posted by RonM:
I propose that runtime be defined as:
The point at which the flashlight's output is one-half of the initial output with new batteries.

<font size="2" face="Verdana, Arial">Ron

So you think all flashaholics who want to test the runtime of their flashlights just to compare it with another one tested by Craig has to buy a luxmeter??? Or what else do you mean with 'output'? For me the runtime always ends when the light output is as low it is no longer useable - and this depends very much on what kind of light (LS or Nichia) and for what I want to use the light.

And don't forget: You then also have to define the initial brightness of each kind of light, the batteries, the room temperature and some other parameters more ....

For me a runtime test will never give me an exact value but more an indication. The efficiency of eg. a DC-DC converter or the light conversion efficiency of a LED are of much more interest and will tell me more of a flashlight than a simple runtime test.

 

[RonM]

flashlights will have a "battery life" that could be measured in SECONDS!!

<font size="2" face="Verdana, Arial">That's why I added the part about initial output actually being measured after the light has been on for 60 seconds.

How is light output to be measured?

<font size="2" face="Verdana, Arial">How you measure doesn't really matter. The only thing that matters is that the measurement be taken the same way and at the same spot in the light beam. Half is half, units are irrelevant.

I'm really serious about establishing this standard, so keep the ideas and critical analysis flowing.

 

[RonM]

For me a runtime test will never give me an exact value but more an indication. The efficiency of eg. a DC-DC converter or the light conversion efficiency of a LED are of much more interest and will tell me more of a flashlight than a simple runtime test.

<font size="2" face="Verdana, Arial">Circuit and LED efficiencies are manifested in flashlight runtime. Lights that are more efficient will produce longer runtimes, plain and simple.

My concern is this...If I choose a light, whether high output or low, I want to know long I can count on it to produce the light that's expected from it. Imagine an LS flashlight that is brighter than 10 Nichias and claims a runtime of 30 hours. Wouldn't it be disappointing if you bought that light and found that after 30 minutes it dims to the level of a 3 Nichia flaslight? That's what I'm getting at.

 

[Brock]

I don't know what a "standard" should be but I just record them in 5 min intervals. I do let them all burn for 1 minute before starting any test to get rid of that battery peak. I will just keep adding to my

http://www.uwgb.edu/nevermab/battery.htm

page. I usually stop recording when they are about 1/6 their origional brightness and I can see them fall off pretty quickly. I am running some top secret test right now, but I just finished my Surefire E22 MN02 runtimes.

 

[brightnorm]

Originally posted by RonM:
I was wondering if we could all agree to a standard of measurement for runtime?....

<font size="2" face="Verdana, Arial">Ron,

I have found Brock's charts (see his link) to be extremely useful and eminently pratical. For example, I wanted to get an LS but I wasn't sure how its brightness/runtime compared to my Inova X5 "Classic". I have the printed chart in front of me and I see that the longest and brightest burning LS is the 2AA with lithiums. At 5:30hrs and a level of "85" it is down 50% from its original brightness. But then I look across the chart at the X5 and see that the LS is down only 13% from the Inova's starting brightness. Then I see that at 5:15hrs the LS is actually 2% brighter than the Inova's initial brightness. Since I have an Inova I know exactly what the relative brightness readings mean and I now have sufficient information to make a decision. I will buy the LS.

But the chart contains many other gold nugget information bits and I'm intrigued by it. I discover things I didn't realize about the AAA, Photon, Opalec, Lightwave 2000 and the Attitude as well as the LS and X5.

I'm amazed by the eight hours of rock steady
"35's" of the MAG/OPALEC conversion, and really surprised at the Photon 3's relative longevity. Since I know exactly what the ARC AAA's starting "17" means in real life terms, I'm able to see and be pleasantly startled by the fact that the LW 2000 matches it at 23:30hrs.

We all get a sense of the capabilities of our lights, and as soon as number/hour values are assigned to those capabilities it becomes very easy to compare them to lights we have never seen that are rated using the same numerical system.

The value of this type of rating sytem cannot be overstated. I realize that making these charts is very time consuming but I hope that Brock continues to expand them because they fill a real need and answer many questions in an eloquently simple and direct manner.

Brightnorm

 

[Chris M.]

Ron have you been reading my review format draft notes?

That`s exactly how I propose to rate runtime of the lights I review. Once, that is, I get things set up to actually take the silly tests!

---
And I Quote:

Runtime Tests
Start after dark! Install new batteries in light. Position light 12" from the sensor and secure both so they cannot move. Turn on light and meter (set to Footcandles and in Record mode so it won`t auto-turn-off), and note the start time. Adjust the light`s position sideways/up/down to give the highest reading from the meter. Close light shielding cover.
Take the peak candela reading after exactly one minute has elapsed, and start auto capture system at the same time. Set to take additional candela readings at sensible intervals (ie if a light will probably run for tens of hours, 15 minute intervals are fine, if a light only goes for an hour or so, minute-ly intervals would be better- use best judgement and go with shorter intervals if uncertain).
Stop the test when the meter reads Zero for two consecutive readings.
Extract and plot the results in a graph according to the blank TRS formatted example in (file path here).
Run time to put in Specs section is the time from start to 50% of the peak.
Odd results? Lather, rinse, repeat (batteries depending). Otherwise note in review.
---

It makes sense to me! Course, I will always still publish the full graph from start to zero, but the "official" runtime figure I deduce from these tests will be the value to 50%. I may also quote time-to-zero, but there`s probably no point as my readers will see that for themselves on the graph in the Long Term Testing section of each completed review.

Now if only I can remember where my light meter is.....

 

[MR Bulk]

Yeah RonM, but -- what if a light is superbright, like say, an UltraStinger or Lambda Hydra or whatever -- when it's only half as bright, it will still blow away many other lights with fresh batteries/charges. So when it's 1/2 depleted it's still more useful, output wise, than those other lights after they've been run just a few minutes. Where's the usefulness in this standard when considered this way?

 

[rlhess]

But Charlie...

If you buy a light because you need 43,500 candela and it drops to 21,750 candela in 3.5 minutes is that any use?

If you buy the light so that the 18,000 candela after 5 hours is useful then the extra light for the first minutes is a bonus.

All of this is approximate as none of us use these tools in exactly the same way.

It's also useful to know run time in some way so that you can plan on how many batteries to take. (and then take more spares).

Cheers,

Richard

 

[Mr Ted Bear]

Guys

I have been doing extensive runtime tests of the "best of the best" lights, LS, Innova, Lightwave, ULTRA, ARC AA etc etc for the past 6 weeks or so.

When you have comparative charts for all the lights (side by side, or actually overlayed on the same chart) there is a wealth of information to be had.

The most important thing that I have discovered about my testing, is that I can not test enough. The scope of my testing needs to be greater, more sampling, and I have recently used over 100 of the 123's just this last month.

Where are my chartss you ask? Soon come at the end of summer.... my charts will be the most comprehensive and extensive on the net.

*entensive meaning several samples of the same light, tested 3-10 times each

 

[RonM]

Some good discussion going on here. I've been away and had limited opportunity to reply.

Chris - Great minds think alike!

Charlie - No doubt, certain lights are many times brighter at 1/2 output than others are at full. My goal isn't to determine which lights are brighter. That's why the unit of measure is irrelevant. The idea is just to determine how long before a light is no longer performing much the way it first did.
Richard - Your point about knowing how to plan for batts is right on the mark.

The absolute best way to evaluate lights is with the types of charts that Brock, Jeff and maybe some others produce. My idea is just for a simple standard that could be used to replace the useless runtimes that the manufacturers state.