Overdriving 5mm leds?

Negeltu

Enlightened
Joined
Aug 28, 2003
Messages
724
Location
Oregon, USA
About a week ago I made a 9 led array on pc board with 5mm 8000mcd leds. They are wired in parallel. I'm running them at about 4.1volts and 30mA of current. It's fairly bright for most purposes. When switching the module to another flashlight body I guess the positive of the battery case made contact ABOVE the resistors and the array got about 4.5 volts from 3 D cells. I couldn't believe how bright it was! I assume they would fry fairly quickly like this? I would appreciate some input on it. How far CAN I push them?
 

shankus

Flashlight Enthusiast
Joined
Feb 16, 2003
Messages
1,472
Location
Mojave, CA
What was the current to them?
My Streamlight 4AA gives about 83 mA to each LED, and last night I was giving my 32 LED module from ebay over 70 mA each LED.
 

Negeltu

Enlightened
Joined
Aug 28, 2003
Messages
724
Location
Oregon, USA
They are getting around 30mA usually. But that one time... it was getting current direct from the 3 D cells and bypassing the resistors because I wasn't paying attention to how the module was situated in the head assembly.
 

Nerd

Flashlight Enthusiast
Joined
Jun 19, 2002
Messages
2,271
Location
Singapore
Run em above 50ma for anything more than 10 mins without good heatsinking and you are asking for trouble. I'm thinking is it feasible to grind off the sides of the epoxy to get to the negative lead and have some metal piece make contact with it to sink away the heat so as to drive them harder.
 

Steelwolf

Flashlight Enthusiast
Joined
Feb 6, 2001
Messages
1,208
Location
Perth, Western Australia
I don't know how bad it would be, but I have a Trek 7, one of the older ones I think. It ran directly from 3 C cells. No resistor anywhere to be seen. Don't use it much anymore because the size to brightness ratio is just pathetic, even after I put in extra LEDs and fresh cells.

I seem to recall, when I first joined these forums, there were quite a few people just banging lights together using 3 cells and a number of LEDs and relying on the internal resistance of the cells to limit the current. I believe one of the key design requirements was to match the number of LEDs to the cell size so as to operate the cells at the point where their internal resistance would start to be felt and distribute the current among an appropriate number of LEDs. Usually 3-6 LEDs for AA, 10-20 for C, 30 or so for D.

I think it was also agreed that overdriving will result in shortened LED life, but so long as you don't get magic smoke straight away, or turn the die black, the LEDs will last long enough that you'll be retiring your flashlight long before you reach the LED half-life anyway. So a little bit of overdriving would be fine.
 

Negeltu

Enlightened
Joined
Aug 28, 2003
Messages
724
Location
Oregon, USA
Thanks alot guys. I switched out some resistors and they are now being driven at about 63mA. I ran the thing for an hour and didn't feel THAT much heat. Seems pretty good to me...and puts out alot of light. I can hit the tops of the trees with it.
 

Rebus

Enlightened
Joined
Sep 15, 2002
Messages
202
Location
Oklahoma
I've had a couple of cheap
multi led direct drive
lights that have leds that
start getting dim after a
few weeks. I know that they
did not have Nichias.
You get what you pay for
with cheap leds.

-Rebus
 

INRETECH

Flashlight Enthusiast
Joined
Aug 22, 2002
Messages
1,318
Location
HILLSBORO, OR
The reason you are not feeling the heat from the 5mm LEDs is that Plastic is a very poor conductor of heat; the only limited heatsinking of the 5mm part is out the leads

My son has one of those white keychains, the the LED "blinks", I suspect that the bond wire is being disconnected from the die of the LED and then reconnected
 

Negeltu

Enlightened
Joined
Aug 28, 2003
Messages
724
Location
Oregon, USA
Mike,

So you think that 63mA will fry them pretty soon? Or just cut their life in half? If I switch from alkaline to NiCd's they should be well within their operating range. Maybe I should just run them on the NiCd's.
 

Negeltu

Enlightened
Joined
Aug 28, 2003
Messages
724
Location
Oregon, USA
I just measured the temp around the leds. When running with 4 AA Alkalines the temp levels off at 107.8 deg Farenheit. When running with 4 AA NiCd's the temp levels off at 94.7 deg Farenheit. I think I'll stick with NiCd's.. /ubbthreads/images/graemlins/wink.gif
 

SockMan

Newly Enlightened
Joined
Nov 12, 2003
Messages
111
Location
Harpursville, NY
I have a Streamlight 4AA. Unfortunately, 3 of the LEDs are now LED "Blinks". The light did tend to get very warm to the point where the silver material on the plastic reflector started to melt.
 

Steelwolf

Flashlight Enthusiast
Joined
Feb 6, 2001
Messages
1,208
Location
Perth, Western Australia
Negeltu: 63mA is quite high. I usually try to limit myself to 40mA. Ensure you have lots of metal in contact with the leg that leds to the die. (Not so important for the leg that leads to the bond wire.)

As for lifetime, there was a guy who figured he used his flashlight an hour a day on average. Not counting vacations when he may leave the light at home, or using a different flashlight, he figured he could get about 54 years out of that light on that usage pattern. And that was only to the LED half-life! /ubbthreads/images/graemlins/eek.gif

I'm not sure about the lifespan for LEDs driven to 60mA though.
 

Negeltu

Enlightened
Joined
Aug 28, 2003
Messages
724
Location
Oregon, USA
Yeah..thanks guys... I took care of that a while back. /ubbthreads/images/graemlins/wink.gif I figured it was too much current. 5mm leds are really cheap, but there's no sense in me needlessly destroying them. I decided to run them at around 40mA. An array of eight 8000mcd whites is plenty bright enough for my purposes.I don't know what it is about 5mm leds, but they are just so addictive.

The Luxeon Stars are a different story. To me they look like works of art.
 

Steelwolf

Flashlight Enthusiast
Joined
Feb 6, 2001
Messages
1,208
Location
Perth, Western Australia
Vince, I thought the 60mA being sent to the Ultra was the maximum point of the pulse? I thought it averaged out to about 40mA?

Negeltu, I know how addictive they are! I just completed building an array of 32 LEDs for a reading light and I haven't even tested out the driver circuit yet! /ubbthreads/images/graemlins/tongue.gif
 

Steelwolf

Flashlight Enthusiast
Joined
Feb 6, 2001
Messages
1,208
Location
Perth, Western Australia
Vince, I'm not sure what you're pointing to as the link brings me to an empty page. But if you're sure that 60mA is the averaged current in, perhaps it's time for me to take out some resistors. /ubbthreads/images/graemlins/smile.gif
 

koala

Flashlight Enthusiast
Joined
Mar 5, 2003
Messages
2,295
Location
Melbourne, Australia
I am not sure why you get an empty page /ubbthreads/images/graemlins/frown.gif but I am pretty sure ARCAAA is also driven at around 60ma avg.

Vince.
 

Doug S

Flashlight Enthusiast
Joined
Jun 20, 2002
Messages
2,712
Location
Chickamauga Georgia
[ QUOTE ]
Steelwolf said:
Vince, I'm not sure what you're pointing to as the link brings me to an empty page. But if you're sure that 60mA is the averaged current in, perhaps it's time for me to take out some resistors. /ubbthreads/images/graemlins/smile.gif

[/ QUOTE ]
Or maybe not. Since heat is the principle concern here, a single 5mm driven at 60mA is a very different situation than a large array of 5mms each driven at 60mA. In the later case it is likely that much higher temperatures will be attained. Somewhat off-topic, the best LED heatsinking that I have seen in a commercially produced single 5mm LED light is in the Dorcy 1AAA that has been the subject of much discussion lately.
 

INRETECH

Flashlight Enthusiast
Joined
Aug 22, 2002
Messages
1,318
Location
HILLSBORO, OR
When the companies print the specifications of 20-30ma, its for a reason; the die in these LEDs are very small and as said before - the only heatsinking is the single lead coming out of the LED that holds the die
 
Top