White LED lumen testing

[jasonck08]

Love this thread! I always find myself coming here looking for data again and again!

I'd like to see some SST-50 tests!

 

[lolzertank]

Do you know how many lux a typical Cree XR-E R2 (any cool white bin will do, but R2 is best) bare emitter has in the center of the beam at some distance? I just got some from a somewhat dubious source, and I want to verify the LED's bin.

EDIT: Never mind, it just occured to me that my lux meter's accuracy is probably so low that I wouldn't be able to tell with any certainty whether my LEDs are a P2 or R2 bin.

 

[metlarules]

Do you know how many lux a typical Cree XR-E R2 (any cool white bin will do, but R2 is best) bare emitter has in the center of the beam at some distance? I just got some from a somewhat dubious source, and I want to verify the LED's bin.

EDIT: Never mind, it just occured to me that my lux meter's accuracy is probably so low that I wouldn't be able to tell with any certainty whether my LEDs are a P2 or R2 bin.

I like your sigline. Its so true.

 

[Erasmus]

Cheers to Jtr1962 for the extensive testing

 

[Nil Einne]

Cree XP-G bin R5 (acquired October 2009)
<snipped>
Overall, the XP-G is another quantum leap in performance from Cree.

Hi, first thanks for all your testing.

But did you forget to include the XP-G R5 data in the zip file? It was last updated in November but I can't seem to see the XP-G data in it and the most recent file by data modified is "Seoul Semiconductor S42180 bin S2 tests.xls" (from October) and I checked, that doesn't seem to be the right thing. Everything else is from September or earlier and nothing there looks right either (although I didn't check each document)

Cheers

 

[jtr1962]

But did you forget to include the XP-G R5 data in the zip file? It was last updated in November but I can't seem to see the XP-G data in it and the most recent file by data modified is "Seoul Semiconductor S42180 bin S2 tests.xls" (from October) and I checked, that doesn't seem to be the right thing. Everything else is from September or earlier and nothing there looks right either (although I didn't check each document)

The strange thing is the S42180 was done the same day as the XP-G R5, and yet the XP-G R5 isn't in the .zip file on my PC, either. Anyway, I just fixed it. The new version of the .zip file now has the XP-G R5.

 

[IMSabbel]

Now that SST-90s arent that exotic anymore, do you have any plans of running one of them as high as it will go?

 

[jtr1962]

Now that SST-90s arent that exotic anymore, do you have any plans of running one of them as high as it will go?

A better choice for that might be the CST-90. Before anyone sends me one though, I need to make and test a higher current power supply. The one I have now ( which I made ) only goes to 11 amps and change. The transformer windings can't take much over that for the periods I would need to test the CST-90. What I want to do then is make a switching supply which can use my existing power supply as the raw voltage. The biggest problem isn't really making a switcher for that kind of current, but rather measuring and controlling the output current to the kind of precision I need for my testing. It'll be done eventually. Once it is, then I'll make a request in this thread for a CST-90. I still have saabluster's SST-90 but it's not a candidate for a high current test since he wants it back. I also have an SST-50 which Fulgeo sent me quite a while back. I have yet to test but will in the very near future. That one is also a loaner, so I need to limit my test currents. Additionally, I have a bunch of small Cree 20 mA LEDs sent to me by JohnR66 which need testing. Now if only I had about 2 days free....

 

[jasonck08]

Any update on the SST-50 tests? Can't wait for the data!

 

[jtr1962]

Any update on the SST-50 tests? Can't wait for the data!

Not yet. To make a long story short I've been going crazy last few months with various issues ( home, business, this latest month long heat wave draining my energy ). Also, I've been putting a lot of cycling miles in last three months and pretty much put my other hobbies to the side for the time being. I think I'm going to try to do the testing this week however. Not much going on business-wise this week, and it looks like we may have a rainy day or two where I can't do any cycling. Besides the SST-50 and Nichia NS6L183, I'll be testing some neutral XP-Gs I purchased this month. Enough testing I'd say to fill the better part of a day.

 

[blasterman]

Sidebar, but Chauvet's Q-spot 260 moving head is powered by a CST-90.

Haven't seen one in action, but reviews put it in the same ballpark as 150watt discharge.

In a compact plastic fixture we can se how jtr1962's tests are so critical.

 

[wquiles]

Besides the SST-50 and Nichia NS6L183, I'll be testing some neutral XP-Gs I purchased this month. Enough testing I'd say to fill the better part of a day.

Which bin of the NS6L183 will you be testing? I used one in this build and I like it a lot.

Will

 

[StupidPig]

Information here is very valuable, thanks.

jtr1962, could you mind to explain a bit on how you convert the lux reading to lumen? I looked into your spreadsheet and it seems very complicated and I'm not quite understand. Thanks.

 

[zzonbi]

Remember 1lux=1lm/m2
So you basically add the lumens back, integrate. Now the finer you go dividing an area, the better. I guess jtr also uses symmetry in led patterns to cut down on sampling.

 

[Vbeez]

Subscribing this. Very valuable info.
Thanks

 

[MikeAusC]

. . . . . . For example, look at this chart for the XP-G:

The red line represents total input power to the LED and the white line represents waste heat. . . . . .

I know this was posted TWO YEARS AGO, but I'm wondering how you determined WASTE HEAT to produce this graph. It's hard to find verifiable information on waste heat from LEDs.

 

[LEDninja]

James Peng,
While the information you provide is very informative, most of us poor CPF members can not afford millions of dollars worth of equipment. We do the best we can with what we have. We do not need 8 decimals of accuracy. Why? because due to manufacturing tolerances a flashlight's tolerances can vary greatly from sample to sample.
Some flashlight manufacturers just quote the biggest number they can find on the LED manufacturers website. As a result almost all cheap flashlights using the SSC-P7 is listed as 900 lumens. When the Cree XM-L came out, to show the superiority of the new LED, these manufacturers claim 1000 lumens (Cree only claim 800 lumens, add loses due to driver and optics and ,,,). When the newer U-bin version of the XM-L came out guess what. They are listed at 1100 lumens. Despite your claim that the integrating sphere in another thread is too small, the reading of ~450 lumens is much more accurate than the numbers I see on ebay.

Another member asked me about a method someone discovered on getting approximate lumens cheaply. Please read it:
http://www.candlepowerforums.com/vb...uring-lumens&p=4322304&highlight=#post4322304
Looking at the numbers of this very crude method I noticed they are within 10% of the lights which are tested by the manufacturers to the FL1 standard.

I used to guess the lumens of my new lights because I have a number of lights of known quantity. I have a 7 LED light of 18 lumens, a Fenix of 29 lumens etc. When I bounce a new light off the ceiling and my reference lights I can tell if my new flashlight is <18 lumens, between 18 lumens and 29 lumens, >29 lumens etc. I was using my room as my 'integrating sphere.

When going out for a walk at night, does it matter if my flashlight is 90 lumens or 91 lumens. I would not be able to tell the difference. Now 200 lumens is definitely brighter. While a national lab needs to be accurate to umpteenth decimal places I do not.

bigchelis, Selfbuilt and JTR1962 are a few who have provided us with reasonably accurate numbers because they have the knowledge and the means to do so.

James Peng, has been banned as a spammer and all his posts deleted. - Norm

 

[Farhaj]

Thatz been a well explained demonstration on the type of LEDs.. If, you can help me with this; it would be Greatful. How much Lumens and Voltage Drops are required for a Generic Chinese 5mm LED to light up completely..

 

[JohnR66]

Thatz been a well explained demonstration on the type of LEDs.. If, you can help me with this; it would be Greatful. How much Lumens and Voltage Drops are required for a Generic Chinese 5mm LED to light up completely..

The typical 5mm LED (white, I'm assuming given the thread title) has a forward voltage (Vf) of around 3.2 volts at 20ma. Generic Chinese 5mm LEDs are rubbish and will fade after a few hundred hours at 20ma. Better to run them at 10ma. If you want something better than the dim bluish light, look for better quality LEDs, such as Cree C503C or C503D series. Much brighter, whiter and last long even at 30ma current.

 

[Farhaj]

The typical 5mm LED (white, I'm assuming given the thread title) has a forward voltage (Vf) of around 3.2 volts at 20ma. Generic Chinese 5mm LEDs are rubbish and will fade after a few hundred hours at 20ma. Better to run them at 10ma. If you want something better than the dim bluish light, look for better quality LEDs, such as Cree C503C or C503D series. Much brighter, whiter and last long even at 30ma current.

I saw your this thread Forum on caculating lumens for white leds... http://www.candlepowerforums.com/vb/showthread.php?89607-White-LED-lumen-testing .....

I have this flash light DP-722b... I want to know its LED Lumens and Watts...

I am sending you its Datasheet... It is 5mm Straw Hat LED... can you please find it out for me the exact Lumens and watts for it..

http://underwaterseaplants.awardspace.com/led.pdf