LED Fade Test Experiment

[kbsvr]

Hi all...

There is a lote of great information in this forum and I would like to contribute something here.

I have done this simple yet somewhat revealing experiment the demonstrates the effect of over powering LEDs even just a little bit. Also it shows the type of power supply drive / limiting methods are effective or destructive to the lamps.

This is a "To be continued..." experiment but at this time may have some useful information for some of us.

The site link is http://www.kbsvr.com/LEDfadetestexperiment/index.html

Feel free to leave any feedback here in this thread.

Kirk

 

[Curious_character]

That's a great experiment. The only problem I have in making use of the data is that there's no indication of how great the light reduction is. Photo-resistors are very nonlinear, so there's no way of knowing, from the data presented, whether the brightness reduction is something so small as to be insignificant or downright huge, or somewhere in between. Some sort of calibrated luminance indication is necessary before we can tell if the results are really important, or just a laboratory curiousity with no practical implications. The output measurements don't have to be absolutely correct (that is, we don't need to know exactly how many lumens or lux are being produced), but they do need to be known relative to each other.

I've gotten a lot of miles from a cheap lux meter I got new for $30 off eBay. I have no way to know its absolute accuracy, but measurements at different distances show it to be linear. I'd think that something like that would be a minimum for obtaining useful results.

c_c

 

[Ganp]

Welcome to CPF Kirk, and thanks for a very interesting post. :goodjob:

I've been wondering how long my direct drive XR-E is going to last :candle:


Colin.

 

[Illum]

well done :twothumbs

say, where do you find a 9VAC transformer anyway?
almost all of these "wall wort" transformers are usually AC-DC stepdown

 

[kbsvr]

Curious_character

Ya I know... this is more of a curiosity experiment with no difinitive results. More of a general figure that illustrates the amount they can fade in a short time. Reason i did this was because i have noticed some flashlight and other automotive light products that don't seem to be so bright after 3 or four years of use and in a few i have disected it turnes out in all cases it has been a state of overdrive to make them meet the desired lumens of output; And pointing out that "Just a little over driven won't hurt" can and does hurt them rather quickly.

I have reasoned for some of them why they are used that way is to obtain high brightness for intermittent useage like a flashlight. Flashlights should not be used as an area light source. Many of them are in overdrive for the brilliance and will not last long. I found this out with my Garity 4 LED flashlight. I have left it on over night several times as a night light and over the past 3 years it has nitocably faded somewhat... i gotta stop! I can talk to much sometimes lol !


Ganp

Hey Thanks! I was directed here by a friend that was doing some research on LED flashlights and he sent me this site. I am happy to be here.

I'm going to be looking for a source on a number of these new LEDs i see people talking about. sounds like a lot of technological improvments have happend in just a year.


Illum_the_nation

Thanks for the comment

USR power supply from an old Sportster 14.4 modem.

 

[paulr]

Hi and welcome, there have been a bunch of experiments like this on different kinds of leds here on CPF in the past, though I don't think anyone has yet tested an XR-E. Check out the "flashlight electronics" section of the forum.

 

[Curious_character]

I have reasoned for some of them why they are used that way is to obtain high brightness for intermittent useage like a flashlight. Flashlights should not be used as an area light source. Many of them are in overdrive for the brilliance and will not last long. I found this out with my Garity 4 LED flashlight. I have left it on over night several times as a night light and over the past 3 years it has nitocably faded somewhat... i gotta stop! I can talk to much sometimes lol !

I have some quantitative data. I have a couple of cheap multi-LED lights with 5 mm LEDs. One is the 9-LED Dorcy that Costco sold a while back at 4/$13 incl. batteries, the other a 21-LED light I picked up at a flea market. both are direct drive, designed for 3 AAA cells. I put an 18650 Li-Ion cell into the 21-LED light and an 18500 into the 9-LED light and ran both through a full discharge cycle. Not long afterward, I measured the light output, and both were just about half what they had been before when running from AAAs.

I've read here that the power LEDs don't seem to deteriorate as badly as 5 mm ones as a result of overdrive, but haven't seen any quantitative data yet. There are a lot of variables -- LED type, temperture, and amount of overdrive for starters. As for me, I'm just a hobbyist, and hardly ever really use a flashlight. And the way things are going, what I have now will probably be replaced by the next generation or modded before any of them gets 10 hours of use.

Again, though, I recommend that anyone even casually interested in things like this shell out 30 bucks or so for a lux meter. With that and a Quickbeam milk-carton light box, you can learn a lot about what your lights are doing.

c_c

 

[2xTrinity]

I've read here that the power LEDs don't seem to deteriorate as badly as 5 mm ones as a result of overdrive, but haven't seen any quantitative data yet. There are a lot of variables -- LED type, temperture, and amount of overdrive for starters. As for me, I'm just a hobbyist, and hardly ever really use a flashlight. And the way things are going, what I have now will probably be replaced by the next generation or modded before any of them gets 10 hours of use.

This suggests that the fundamental variable for overdrive is heat buildup. In a power LED, you're likely going to be running them connected to some sort of heat sink. In the case of an encapsulated 5mm LED, they will tend to retain a lot of the heat and thus be more prone to reach excessive temepratures if overdriven.

Also, since they are lower output to begin with, peopel will also be more tempted to overdrive them in the first place compared to something like a Cree LED where even 1/3rd of the rated power is a lot of light.

 

[AndyTiedye]

How effective is the light meter on a good camera for doing such comparisons?

 

[Curious_character]

How effective is the light meter on a good camera for doing such comparisons?

I'd expect it to be pretty good, although I don't have any experience one way or the other. You'd have to deal with the scale, and such things as the square law relationship of F stop to light level. You should even be able to come up with absolute lux values if you want -- I'm sure that conversion information is available somewhere on the web.

c_c