White LED lumen testing

[ViReN]

jtr1962 .... Great plots in deed... its valuable data when it comes to designing a light considering the Lumen/Watt & total Lumen output for a 5 mm LED

I wonder if you could help me for a project, if you have any Luxeon I (Q/R Bin's H/J Vf) and Luxeon III (S/T Bin's J/K Vf) LED's is it possible for you to have similar charts. i.e. Lumens V/S Current & Vf

i have been looking for this over the forum here and there... but could not find such a chart. i do not have the equipment to do the testing either

your help is highly appreciated.

 

[jtr1962]

I wonder if you could help me for a project, if you have any Luxeon I (Q/R Bin's H/J Vf) and Luxeon III (S/T Bin's J/K Vf) LED's is it possible for you to have similar charts. i.e. Lumens V/S Current & Vf

Hi Viren,

Thanks for the compliments, and I hope this data is useful to people. If you look a few posts up you'll see that I did indeed plot a Luxeon Q3J in the same manner as my 5mm LEDs, and included a link for the relevant spreadsheets which have the Vf versus current data. I don't have any Luxeon IIIs on hand to do plots, but I imagine they wouldn't be much different. Maybe now that Future Electronics finally lowered their prices for these to something I consider reasonable I'll pick up a few LIIIs in various colors and play around.

 

[winny]

Good job jr1962!

I didn't read it all but as you used the word steradian so I'm quite confident that you know how it works. Very good. Welcome to the higher intelligence club! ;-p

From the picture it looked like you where using the same light meter as I am. I'll give you a warning on them. They are cheap and in my manual it said "the spectral response almost corresponds with the CIE curve". When you measure non 2860 K light sources, like LEDs, the error margin is big and comparing LEDs with different shades of white or with incans is, ..., questionable.

 

[jtr1962]

From the picture it looked like you where using the same light meter as I am. I'll give you a warning on them. They are cheap and in my manual it said "the spectral response almost corresponds with the CIE curve". When you measure non 2860 K light sources, like LEDs, the error margin is big and comparing LEDs with different shades of white or with incans is well, questionable.

The light meter is a CEM DT-1300. I took part in the light meter tests as described here. For LED-type white light my meter only seems to read a few percent above the averages of everyone elses. It does read quite high for blue LED light which explains my ridiculously high results in tests of blue LEDs (about twice as high as they should be).

Note that I did sanity check my results. The brightest 5mm LEDs nowadays are in the 60 to 70 lm/W range so if I had gotten results much different I would have been suspicious. Also, my Q bin Luxeon came in at 37 lumens, solidly in the middle of the Q bin. Again, if I had gotten results of 45 or 50 lumens, I might have been suspicious. Overall, I'd say for white LEDs my results might be about 5% or so high, if that. I'm more interested in relative measurements anyway so as to tell which LEDs are better. In the future I may adjust my data once the CPF standard LED test lights are professionally checked but for now the relative measurements are a good guide for everyone.

 

[winny]

You obviously have thought of everything! Very good.

I hope I didn't offend you, but I just had to warn you as my light meter can be waaay of when it comes to LEDs sometimes. I'll send mine to SilverFox for testing when he get the lamps back.

Anyway, how much did you get out of that MR16 in the background of the picture?

 

[jtr1962]

Anyway, how much did you get out of that MR16 in the background of the picture?

I never tested it but now that you mentioned it I think I will. BTW, it's a standard 50W halogen MR-16 so I figure it should be good for ~1000 lumens.

 

[jtr1962]

I added the Jeled 50,000 mcd to the list in the first post and updated the graphs accordingly. You might need to refresh your browser to see the updated graphs. The relevant spreadsheets were added to the .zip file linked to in the first post of this thread.

 

[winny]

Any progress on the MR16?

 

[jtr1962]

Any progress on the MR16?

Unfortunately, that picture was taken before I cleaned the room and now I can't remember where I put the MR16. :green: If/when it turns up I'll be sure to test it and post the results. I did actually look for it for a while when you first mentioned testing it but had no luck.

 

[Pinter]

For LED-type white light my meter only seems to read a few percent above the averages of everyone elses. It does read quite high for blue LED light which explains my ridiculously high results in tests of blue LEDs (about twice as high as they should be).

That means that leds with blue hotspot are resulting in higher Lux reading in your test than real white ones?

 

[hotbeam]

jtr, fantastic set of results. you must be a very persistent fellow

 

[jtr1962]

That means that leds with blue hotspot are resulting in higher Lux reading in your test than real white ones?

Maybe except that the greatest errors with my meter occur with narrowband blue light such as from a blue LED, rather than the bluish-white hotspots present in some white LEDs. As a reality check on the accuracy of my light meter I compare my intensity results to ratings. If anything, most eBay sellers overrate their LEDs so I should be getting lower intensities. In fact, that is almost always the case. Only a few sellers, such as BestHongKong and ChiWing, usually live up to their ratings, or occasionally even exceed them. The only time I get intensity readings well above ratings across many manufacturers is with blue LEDs. The CPF light meter test verified that my meter is indeed reading very high for blue light but pretty much on the mark for white light (both incandescent and LED).

 

[jtr1962]

jtr, fantastic set of results. you must be a very persistent fellow

Obsessive-compulsive is sometimes a better description of me.

 

[vortechs]

Hi jtr1962,

I was really happy to find this thread reviewing the 50k white LED's from jeled. Thanks for all your work to compare the real statistics and output from all of these different LED types.

Are you still running new tests on white LED's and, if so, what LED's are you planning to test next?

I've been thinking about getting the 3mm white 18,000mcd LED's available from jeledhk on eBay (link). Their published stats show 15,000 to 18,000mcd with a 30 degree viewing angle, but I'm wondering how close to this they actually are. I've been considering getting some to modify one of KevinL's picolights that use 3mm LED's and possibly for a solitare or 2AA maglite mod without needing the reflector modification that using a 5mm LED would require.

If you've already tested these 18k 3mm white LED's or are planning to test them, please let me know. Alternatively, if you are interested, I could get some and send you a few to test (I don't know what I'd do with 50 of them anyway).

 

[jtr1962]

Are you still running new tests on white LED's and, if so, what LED's are you planning to test next?

These tests are ongoing, and I just updated the first post with the Peak Snow 29 and Nichia CS B1U LEDs sent to me by CPF member Pinter. I'm accepting LEDs for testing, and also will continue testing any new ones I acquire for my own personal use.

If you've already tested these 18k 3mm white LED's or are planning to test them, please let me know. Alternatively, if you are interested, I could get some and send you a few to test (I don't know what I'd do with 50 of them anyway).

If you've gotten the 18K 3mm LEDs and have some left I'll be happy to test them. Just PM me for a mailing address.

 

[jtr1962]

I also tested the Seoul Semiconductor W32182 1 watt stars. The results are in this post. Interestingly, I got an luminous flux of almost exactly the 52 lumens claimed as typical at 350 mA for this part.

 

[winny]

Have you found that MR16 yet?

 

[jtr1962]

Have you found that MR16 yet?

Not yet, but I think I have a good idea where I put it. Stay tuned....

 

[Pinter]

Hi jtr1962,

Thank you for testing the Nichias and Peak Snow, I did not think you measure all 9 samples.
Nichia performance was a real surprise over 60mA. This is in good correlation with another thread in which Nichia has shown better lumen maintenance at 60mA than others at 30mA.

I have some questions about the methodology:
- what is the sensor diameter
- (whether if I understood correctly) you measure the beam profile from 17.5" then you read the peak lux readings from 1m on different currents and weight this data with the relative intensity that came from the profile?

 

[jtr1962]

I have some questions about the methodology:
- what is the sensor diameter
- (whether if I understood correctly) you measure the beam profile from 17.5" then you read the peak lux readings from 1m on different currents and weight this data with the relative intensity that came from the profile?

Sensor diameter is 1.5". I chose 17.5" as the measuring distance because the sensor diameter at that distance covers exactly 5° of the beam, and I measure output in 5° increments. When I'm done I weight the beam profile data with my lux readings from exactly one meter, and use a spreadsheet to do the calculations. The end result is of course the total output in lumens. Yes, I do all the lux readings at different currents at one meter, although since they're all relative anyway I could just as easily do them at 17.5", and only do the 1 meter lux reading at one current just to weight all my data. The method works surprising well. The main source of error seems to be the absolute value of my lux readings at 1 meter. The meter is supposed to be within 10%, so I assume this is the upper bounds of my error. In practice, based on my relative measurements in Silverfox's light meter benchmark testing, my light meter is reading maybe 5% high for white LED type light compared to the average (although the average isn't necessarily dead accurate).

In any case, my results aren't horribly off. The U bin Nichias are supposed to be in the vicinity of 60 lm/W efficiency at 20 mA but Vf is typically about 3.4 volts. The samples you sent me had a lower Vf in the 3.2V area which would put efficiency into the mid 60s. Furthermore, all seemed to be at the upper end of the U bin, probably pushing them into the 70 to 75 lm/W area. Since 75 lm/W is what I measured, my errors are likely 5% or less. This also explains why V bin Nichias are very rare. They need to have efficiencies of 80 lm/W or higher, and this represents the present upper limit of production white LEDs these days.