Light Meter Benchmark Testing – CPF style

Light Meter Benchmark Testing – CPF style 10/2/05

Have you ever wondered about the accuracy of your light meter?

Have you ever wondered why someone else seems to report higher (or lower) lux values than you observe, even when you are measuring the same type of light?

What if we could take one light and have everyone measure the lux at 1 meter or foot candles at 1 foot and log in their findings?

I want to open this topic up for discussion. Please understand that there are no "right" answers. You measure what your light meter says and report it here. All I am trying to do is level the playing field by providing stable light sources for the comparison.

All discussion will be conducted in this thread. The pass around thread will only contain those who want to participate. Discuss the method and report the findings here in this thread.

Don (McGizmo) will provide an Aleph light based on the NextGen driver running a Lux III LED at 500 mA. He also will be providing a beam shaper that can be used for another optional measurement if you want to do that. In addition, Don will be sending another head along that has a blue K2 LED in it. It will fit the same battery holder as the LUX III.

Note: After the first round of testing by Don and some additional testing by me, Don is adding Red and Green LED lights to the package. This will give us three points of the spectrum (red, green and blue) as well as white LED and an incandescent.

PK of SureFire will be providing an A2 without LED's for this testing.

These three lights will become the CPF light standards. They will be used only for testing. We are trying to conserve the LED and Lamp life.

To participate in this round robin testing:

1. - You will need to have a light meter (be sure to check the battery).

2. - You will need to have an accurate way to measure 1 meter (which is 39 3/8"). If you measure and report in foot candles, your measurement should be taken at 1 foot. If your meter measures both lux and foot candles, I will leave it up to you to decide if you want to report both.

3. - You will need to have three (3) fresh primary CR123A batteries (any brand is OK) one for the Aleph and two for the A2 – NO RECHARGEABLE CELLS will be used during this testing. The LUX III will be tested first, then the blue K2 will be tested on the same battery.

Note: We have just added a red and green LED light to be tested. You should be able to test the white, blue, green, and red LED lights on one battery (in that order), but if you find that it is taking a bit longer than you had planned, you can use the batteries from the A2 (after you have tested the A2 of course).

4. - You will need to take the measurements as you would normally do. We hope that you have a dark area that is as free from reflections as possible, but please report your findings as you would normally do the measurements. No special set up is required. You may do another set of measurements if you want to tidy things up and report that as well.

5 – You will be required to follow the "CPF Standard Procedure for Light Meter Measurements" listed below.

6 - You will need to post what your meter reads, along with the brand and model of meter. We will also need the room temperature during the measurements.

7. – You will need to contact the next person in line to make sure they are available to participate before sending the light to them.

8 - You will need to pay the postage and insurance required (the insurance value of these lights is $500, and we can call them benchmarking lamps) to send it to the next person and make sure it gets packaged securely and sent out.

At the end of the testing, the lights will be returned to me. I will store the lights at the end of the testing. They will be the CPF standard light sources. The will be available for future tests by arrangement. These "standards" will not be used for general illumination, but will be reserved only for testing purposes.

CPF Standard Procedure for Light Meter Measurements

1. - You can not alter, modify, or take apart these lights. Everything needs to be aligned and kept in alignment so we have a good basis for comparison. Opening the bezel could allow the reflector to shift and that would alter the following measurements.

2. - Your light meter will be set up with 1 meter distance for those measuring in lux (or 1 foot for those measuring in foot candles) between the dome of the meter sensor and the outside edge of the bezel ring,

3. - Check the lens to make sure it is clean and free from fingerprints, and make sure the battery in your light meter is good.

4. - The lights will be turned on for a minimum of two minutes before taking measurements. This goes for all lights.

5. - You should take the measurements as you normally would and report what you get. You can then do another set of measurements under tighter control and report those measurements as well if you are so inclined to do so.

6. – You should take a minimum of three measurements to check the repeatability of your set up. This means taking your system apart and putting it back together each time. We are checking to see if you can repeat the same numbers each time you set up. Report your three readings so we can see how much scatter there is, then average them together for your overall number.

Schedule

Lets try to keep things moving along and make every effort to get the lights sent to the next person within 2 days.

Please understand that these lights are intended for testing purposes only. They have a limited life time and should not be used for general illumination needs, or battery run time tests.

At the end of the pass around (or at the end of the US portion depending on what we can set up with the laboratory), the lights will be sent to a professional laboratory along with a copy of our "standard procedure." Their results will allow us to "adjust" our readings as necessary to better reflect reality.

It should only take a few minutes to take your measurements, post them, contact the next person in line, and pack the light up to send off. I know that unexpected things come up but let's try to keep things going and make sure it gets sent to the next person within 2 days.

I think this will go real quick and believe it would be a great service to the CPF community.

What do you think? Should I set it up? It is in the works...

Tom

You can sign up for the pass around here.

Here is the list of people participating in this testing.

In the US
McGizmo - finished testing. Results posted here.
SilverFox - finished
The LED Museum - finished
Bwaites - finished
Sigman - finished
Bogus1 - finished
Evan9162 - finished
Chevrofreak - finished
Andrewwynn - finished
N162E - finished
Wptski - finished
JTR1962 - finished
Rdshores - Passed Away 12/31/05 - RIP
Quickbeam - finished
Jtice - finished
Kitelights - finished
4sevens - finished
Wquiles - finished
Cy - finished
Archangel - finished
Brock - finished but need meter model and serial number
Codeman - finished
Off to Lighting Science Inc for official certification - finished
Lildave - finished
Yaesumofo - finished
Gadget Lover - finished
Bullzeyebill - finished
Geroges80 - finished
NextLight
Robstarr-Lite

Tvodrd
HarryN - on hold until later
Modamag

NewBie - with a Minolta CS-1000
McGizmo

In Puerto Rico
Reima

In Canada
We did not have a lot of response from Canada, so Chimo who has a Meterman decided to pass on this round. I had suggested he send his meter to someone in the US so he could avoid the extra taxes and duties involved in shipping the test lights to Canada. Perhaps we can get the serial number of his Meterman to see where it fits in with the other meters...

In Europe
Kiessling
Mark2
Winny

In Argentina
PEU

Tom

The results are very interesting…



What is also interesting is looking at the standard deviation we are getting.




The official results from LSI are:
A2 Incandescent - 2508 lux
White LED - 1270 lux
Blue LED - 151 lux
Green LED - 2357 lux
Red LED - 2379 lux

Awesome. I like the idea. We'll have a good standard of measurement for comparison. (and we can 'calibrate' our meters to the CPF gold standard Aleph)

One thing about the measurement technique: After turning on the light, the Lux will warm up a bit. Are we to take a measurement immediately, or is there some time we should wait for the light to reach thermal equilibrium?

I'd be interested simply to see how valid my piecewise lumen integration technique is.

I'm in. Do you think that different battery brands will show different results. Like some of the runtime testing of say even the copper top Duracell and the Duracell Ultra. (IIRC) Would they have slightly different voltage output and skew the results?

I'm in, great idea!

Tom:

What would be the procedure for locating the light sensor in the head's beam or spot? Centered for highest reading, I assume?

Just want to throw out that "centered" and "highest reading" aren't always the same place.

And that i'm interested in getting in on this.

wptski said:

What would be the procedure for locating the light sensor in the head's beam or spot? Centered for highest reading, I assume?

Click to expand...

Remember:

1 Fc@1Ft = 1Cd(candella) = 1Lux@1M

Hello Evan,

Your are the other person I had in mind when I mentioned JTR, but I couldn't remember who you were. Thanks for jumping in here. I was going to do a search, but am not sure the search function is fully functional yet.

Don is going to send me a "new" light. I will run it for 4-5 hours to break it in before starting the testing. Does that seem about right?

My thought is to turn the light on and take the measurement after 1 minute of warm up. Does that sound OK?

Tom

Great Idea !!! I'm in

Hello CromagNet,

The NextGen driver will regulate the current to the LED so battery voltage does not really matter.

I discussed this at some length with Don and we decided that the type of battery will not matter for this test, as long as it has a high enough voltage to bring the light into regulation.

Total run time is a different test that is dependent on battery brand, but every battery I have seen has been able to handle 500 mA for at least a few minutes.

Tom

Hello Bill,

Yes, we are looking for the highest reading. Some meters have a max hold or peak hold that makes that a bit easier to read.

Tom

Hello Roy,

That is correct.

I have seen some light/reflector combinations that influence closer readings. I don't think it would be a problem with the Aleph 2, but in the interest of having everyone measure the same beam, I suggested taking a fc measurement at 1 yard and working backwards.

When I receive the light, I will do a measurement at 1 foot and see if there is any reason for concern. As you have pointed out, the fc measurement at 1 foot should equal the lux measurement taken at 1 meter.

Tom

Tom,


4-5 hours wouldn't be a bad idea. 24 would be more thorough, but overkill in this case since it's got a good current regulator on board.

As for testing time, since (i'm assuming) you'll have first shot at this, why not during your break-in, set up your light meter and take measurements throughout the run - turn-on, 1 minute, 5 minutes, 20 minutes, 1 hour...Then maybe repeat the first 4 measurements after the 5 hour test to verify - then you'll know when the reading will level out, and how long the light will need to be turned on before getting a "good" reading.

Maybe don can tell us about how long the aleph takes to reach thermal equilibrium and just use that value?

At 500mA, though, I can't see heating up making more than about a 5% difference. Good idea on using a relatively "low" powered light, it makes for a more stable source.

Hello Evan,

Good ideas...

To break the light in, I also have the option to hook it up to my power supply and let it run for a day or two.

On the other hand, for our actual measurement, does the actual time really matter? If we are all taking the measurement at the same elapsed time, we should all be at the same thermal gradient level.

Tom

This is an EXCELLENT idea !!!

I would gladly take a turn taking a measurement.

This has been a problem for a long time,
everyones conditions are very different, and different meters will take a wide range of measurements.

One thing I have noticed,
my PEAK Lux readings are very different from what I can "peak" at manually.
On average, MY peak reading is higher than the meters peak reading.
Anyone else notice that?

Also, we are gonna have a problem here,
even though you are having everyone use a "fresh" cell,
they WILL be different, in both voltage, and resistance.
I am not sure how much that will effect the regulated Nexgen though.

[EDIT] eh, just reread some above posts.
Seems Don says the batteries shouldnt matter, as long as they are up to the proper voltage.
I hope that is true.

~John

I'm in.

This will complicate things, but what about an incandescent source for the test?

One of each type of light might be useful!

Bill

Hi guys,

I have a UX1J on an E-screw currently runing on a bench supply at 500 mA to break in the LED. I'll let it run for over 24 hours straight before I build out the light engine. This should bring the die well into its lumen maintenance curve and the elapsed time used in the various meter tests shouldn't have a significant impact on the LED. I will measure it before sending to Tom and then I can re measure it upon its return. I have noticed that the NexGen converters seem to be rock steady in terms of current control and probably 10 converters set at 500 that I actually measured all showed 503 mA current actually going to the LED. This drive level on the LuxIII and especially with an efficient UX1J should not produce levels ot thermal output that would skew our readings, IMHO.

This is not a contest to see who can get the highest reading!! The idea here is for all of us to get a sense of relative measures and how our meters and methods compare to other meters. It will also be interesting to see how much variation we see in measurements of the same light. YMMV is always at play and this will give us an idea of the degree to which YMMV based on presumably the same light source.

I was thinking a measure with the beam shaper as well as the reflector might give us more data as the beam shapper provides a more homogeneous distribution of light and the lux varies much less across the beam. It also gives us two output levels to measure and report.

EDIT: This light will belong to CPF and it should be available for future tests and measurements if there is a "keeper" of the light willing to take on the responsibility (Tom??? ) It might also be worth considering having this light measured by a professional firm for both lux as well as flux. :shrug:

You're probably right, it probably won't make too much of a difference.

[rambling]
Though the environment can make a difference. You will see a 5% difference, if say, I do the measurements in my garage at 10C, and someone else does their measurements in the house at 25C. Or, if someone does measurements in their garage in winter at 10C, then 6 months later (shouldn't take that long, but just an example), it gets to someone who tests in their garage where it's 30C, then you could see a 7% difference.
You get 5% difference for every 15C difference in junction temp, which also means you get 5% difference for every 15C difference in ambient temp.

[/rambling]

I'm probably just being too picky in thinking about controlling for every variable we can and just having the light meter be the variable under test.

It'd probably be a good thing to describe the test environment, then we can sort of control for that after the fact.

So, after all that, measuring at 1 minute sounds good to me :thumbsup:

The Beam shaper is a very good idea Don,
REAL good idea actually.

I think everyone that has a meter that can do Peak readings,
should take one with the peak set to on, AND a reading seeing what peak they can get manually.

And yes, each person should describe their test environment also.
Along with what meter they have of corse.

~John

For an incandescent source, it should be a SF A2 where we have voltage regulation, yes? Perhaps one with the LED ring removed as the LED's are a function of Vin.

EDIT: I will be installing a single stage switch on this light so you will only have regulated high to measure. THe resistored low is a function of the battery and not likely something we should consider "controlled".

I would also like to point out that the variations of measure due to ambient conditions are at play every time someone reports a meter reading and it is most important for the measurer to be consistant in their method and environment; not only in measuring this light but in other measures given on other lights. :shrug: