5mm LED fade test. JELEDs, LOV, RS

[JohnR66]

I have another "fade" test completed. This time I tested a Radio Shack, the popular JELED 55cd white and Light of Victory 55cd white LEDs. Each LED was fed a current of 30ma for 8 days continuous (192 hours). I used my digital SLR camera to photograph the LEDs through a piece of white plexi sitting directly on top of the LEDs (0 distance). The Left column is the tested group, the Right LEDs are the control group that were not driven during the test. So here we go...

1st Row: Radio Shack 276-0017 LED. These are not really warm white, but more closer to 4,000 Kelvin color temp. They are bright for their color temp. Amazing is they did not fade one bit. It was the brightest LED at the end of the test. 30 Deg beam is good for low current fixed lighting or floody flashlight use. Beam is pretty consistant for a 5mm LED. Great product :thumbsup:

2nd Row: Ebay JELED 55cd white. I was hoping for good performance from this bright, narrow beam LED. I was let down in that it faded this much in the test. It is still okay for flashlight use, but I'd recommend not pushing them too hard.

3rd Row: Ebay Light of Victory 55cd white. It too showed some fading. I tested them before and thought they did better in the first round. These have a wider beam than the JELED and a consistant color.

Conclusion: This Radio Shack LED is a quality LED. At $2 for a two pack, they aren't cheap, but many pay more for Nichias. I like the warmish neutral white color and their performance in the test make them a recommended product.

 

[Fallingwater]

Bugger. I was hoping for the Jeleds to work better.

 

[EngrPaul]

Thanks for your time and the info! :twothumbs

 

[PCC]

That's not too bad. Could have been worse, I guess.

 

[jtr1962]

These results aren't as bad as they seem. Don't forget that the lifespan is an exponential function of current. At 20 mA they might last 5 or 10 times as long. At 10 mA they'll probably last 50 times as long or more. If lifetime is any concern, then 5mm LEDs shouldn't be driven at over 10 mA (and 5 mA is even better).

 

[EngrPaul]

I put JELED's in a nitelight at 20mA, and they went dark in a month.

Later replaced with a luxeon take-off emitters, and they are lasting years.

 

[JohnR66]

jtr1962: My LOVs faded about the same after a month at 15ma. That is 4 times longer, so it is exponential, but life it still too short.

EngrPaul: What was the current crest factor of the night light (assuming AC to DC driver) How old were the JELEDs? You said the new Luxeon have worked for years. Could the JELEDs be older/different product?

 

[Black Rose]

I may have missed it, but have any of the CREE 5mm LEDs been tested?

 

[EngrPaul]

EngrPaul: What was the current crest factor of the night light (assuming AC to DC driver) How old were the JELEDs? You said the new Luxeon have worked for years. Could the JELEDs be older/different product?

It was a while ago that I bought the JELED's. The current was RMS on a DMM.

 

[JohnR66]

I may have missed it, but have any of the CREE 5mm LEDs been tested?

I have not tested them. I'd love to test them and some Nichia LEDs. Until work picks up, I won't be able to buy more LEDs.

I have some warm white LOV 20ma flux LEDs on the way. Ordered a couple weeks ago.

 

[Phaserburn]

Great tests! I agree with 30ma or even more as the drive current, because that's what they are almost always driven at in flashlights, lanterns, nightlights, etc. Quite often, they are driven much higher.

 

[Fallingwater]

I have not tested them. I'd love to test them and some Nichia LEDs. Until work picks up, I won't be able to buy more LEDs.

I have some warm white LOV 20ma flux LEDs on the way. Ordered a couple weeks ago.

User Buck91 was giving away some leftover unused Nichia GS he had. Try contacting him, he might be willing to donate one or two for the good of mankind.

I suspect that it's the phosphor, rather than the actual light-emitting die, that's being cooked in these LEDs. Say, how about doing fade tests of blue LEDs? I think they'd last rather longer.

 

[RyanA]

I wonder if the LEDs would last longer if they were thermally conductive. I don't know what the material that is used to encase the die would be but perhaps if they could get something like aluminum oxide into the mix near the base of the bulb somehow.:shrug:

 

[EngrPaul]

1st Row: Radio Shack 276-0017 LED. These are not really warm white, but more closer to 4,000 Kelvin color temp. They are bright for their color temp. Amazing is they did not fade one bit. It was the brightest LED at the end of the test. 30 Deg beam is good for low current fixed lighting or floody flashlight use. Beam is pretty consistant for a 5mm LED. Great product :thumbsup:

I bought some over at my local RS, and found the same thing. Not warm white, but definitely not cold either! No tint variations, not sure how much is due to the high quality die, or from the wide viewing angle.

I'll try modifying one of my purplish lights with these and see how it goes.

 

[ledstein]

Fate test with a digital camera is not very precise. At least use a luxmeter.

For usual LEDs (China) 15mA is the max if you care at all about lifetime.

 

[JohnR66]

Fate test with a digital camera is not very precise. At least use a luxmeter.

For usual LEDs (China) 15mA is the max if you care at all about lifetime.

I disagree. While a lux meter will show the numbers, a camera shows if a LED has lost light output quite clearly. Opening the file in a image editing program and checking the RGB values can show very slight differences that are not visible. The integrating sphere would be the ideal tool. For me, I just need to see how they hold up and my test shows go-no go clearly.

15ma is too much for some if you want them to last over a month. The LOV's faded after a month. The JELED is probably the same given the results of this 1 week test. The Rat Shack LEDs didn't fade one bit after 1 month at 15ma but had a very slight warmer color shift.

 

[Phaserburn]

I wonder if the RS leds are Nichias.

:thinking:

 

[Fallingwater]

If the camera is used with the exact same manual settings in the same position, I don't see why it can't be used to measure fading.

 

[EngrPaul]

You'd have to put the fading in a graphical format anyway, most of us can't comprehend magnitude of numbers without pictures. :nana:

 

[EngrPaul]

It's important to note that many of these LED's don't use tapered lenses, and therefore are too big for the mounting hole. They won't drop into many applications (photons, fenix, etc.) without reaming. The Radio Shack LED is one, and the JELED's I had a while back were the same.