LED Heat Question

[cmaylodm]

I recently modified an Ultrafire WF-602A1 with a Seoul P4 (USWOI). Before, with the stock LED, the body of the flashlight heated up but never got uncomfortably hot even when I set it on the desk and ran it for ~30 minutes. When I replaced the stock LED with the Seoul, I used Arctic Alumina Expoy to attach the LED to the slug instead of the crappy generic thermal goop that was on there. Now when I run the light for ~30 minutes sitting on my desk, it gets very hot, almost like an incandescent. The emitted light didn't change color or intensity much during the change from cold to hot, and I haven't detected any short circuits on the pill. Is the change in heat just because I am using much better thermal paste, thus achieving much better thermal conductivity?

 

[AlexDeRus]

what thermal paste did you use exactly? what brand is it? where can it be purchased?

 

[eebowler]

cmaylodm I would assume that is the reason as well. Be aware too that the new LED may have a lower foward voltage than the previous one and since the A1 is likely voltage regulated, a lower Vf would result in a higher current draw from the LED.

 

[Illum]

Arctic Alumnia is essentially thermal paste in epoxy form, so using both together without cleaning one off your end up with a mess that doesnt stick.

Arctic Alumnia is probably the best thermal transfer compound I know...but unfortunately its epoxy so its not removeable

if your feeling a significantly increasing temperature, you know the termal compund is working, also...check your resistor, you might be overdriving the led, giving it more excitement than it bargained for....30 minutes is a moderately long runtime, it should feel warm, but not hot to the touch

hope this helps, cheers

 

[mzzj]

Is the change in heat just because I am using much better thermal paste, thus achieving much better thermal conductivity?

Can't be, because only exit for heat is still trough lamp body. Mass of led/star is so low that it can't store even a fraction of the thermal energy generated in 30mins without heating to smoking hot 900F.

 

[cmaylodm]

AlexDeRus - Arctic Alumina Thermal Epoxy, purchased from newegg.com

eebowler - Using my multimeter to test current, I noticed it shift from 350mA with the old Luxeon to 500mA with the Seoul on a freshly charged 14500. By putting in an LED with a lower Vf, does the driver increase current until the Vf matches the voltage specified by the regulator?

Illum_the_nation - The body of the flashlight gets really hot (like the bezel of a 6v D26 incan after its been on for ~10 minutes) using the Seoul attached with Arctic alumina when it's sitting on my desk running. Holding onto it with my hand, it gets warm, but never hot. I don't smell any magic smoke from the driver, and the battery isn't hot, so could I conclude that this heat is good heat? Sorry for all the questions, I am new to the solid-state lumen game and never really knew what kind of heat to expect from an LED.

 

[65535]

with 500mA and a rather small light, plus improved thermal coductivity, it doesn't surprise me its getting nice and toasty, really these lights aren't meant for use just sitting in a warm room on a desk, in your hand you will absorb the heat fast enough and well enough I doubt you will realize any change in temp, your LED is working well, at 500mA their is a good deal of heat, so nothing to worry about. Try to keep it cool though.

 

[eebowler]

eebowler - Using my multimeter to test current, I noticed it shift from 350mA with the old Luxeon to 500mA with the Seoul on a freshly charged 14500. By putting in an LED with a lower Vf, does the driver increase current until the Vf matches the voltage specified by the regulator?

For an IDEAL voltage regulator yes however, the average voltage 'regulator' is actually a voltage booster with some regulation. In effect, outut voltage is determined by the input voltage but isn't proportional to it. For eg, A 1.5V input may result in a current of 200mA through the LED while a 3V input may result in 300mA to the LED.

In your situation, because the input voltage is the same in both cases, the output voltage from the regulator would be the same also and that voltage across the original LED results in a current of 350mA flowing while through the seoul LED 500mA flows.

Makes sense?

It's my opinion that the heat is good heat.

 

[cratz2]

Hrmm... Well, without measuring AT the LED, it's hard to know exactly what is going on. But overall, I'd have to agree with 'the heat is good' crew. I've never modded that particular light and know basically nothing about it, but I have to assume the original stock light was developing heat. Heck, they say that an SMJLED in a PR2 base can get to... what was it... 140 degrees Fahrenheit right at the emitter? I'm sure a Lux at 200 to 300 develops some kind of heat and if you can't feel it at all through the body, then something is wrong. Remember, air is a VERY good thermal isolater.

Personally, I'd keep the modded light and just and use it normally... and running it 30 minutes nonstop without touching it isn't really all that normal... outside of CPF circles, that is.

Do you know the Vf values of the original and replacement stars or were they emitters?

 

[cmaylodm]

I don't know about the stock LED that was in there, but I measured a Vf of 3.6v for the Seoul. Considering that its binned with a Vf of 3.25-3.5 at 350mA, the shift to 3.6v with a little more current makes sense to me. IIRC, NewBie had a graph showing the Vf increasing with increasing current, up to ~3.8v at 1A.

Hopefully today I'll get my FLuPIC and then the real testing will start :rock:

 

[Illum]

with 500mA and a rather small light, plus improved thermal coductivity, it doesn't surprise me its getting nice and toasty, really these lights aren't meant for use just sitting in a warm room on a desk, in your hand you will absorb the heat fast enough and well enough I doubt you will realize any change in temp, your LED is working well, at 500mA their is a good deal of heat, so nothing to worry about. Try to keep it cool though.

good point, Im pretty sure the light can stand alone with no problems..as long as your not detecting the same temperature increase in the battery

 

[yaesumofo]

My experience with theses Seoul emitters is that they run much cooler than their Luxeon counterparts.

In fact a Seoul when driven at the same level as a lux III takes a very long time only to warm up a few degrees.
I have an ARC LS with a Cree driven at 525 and this light doesn't get hot at all. When a Luxeon emitter was used in this light it got VERY hot.
I have a PD-S driven at 390 ma with zero heating.
My point is that if you are getting a very hot system, I would guess that the emitter is being driven at or very close to it's maximum.
Be careful.
Yaesumofo