Re: about the thermal resistance of LEDs.



On Sep 17, 11:31 am, "Raymond Chou" <93246...@xxxxxxxxxxxxx> wrote:
Hi all:
I have a question about thermal resistance of LEDs.
For a single LED, is the thermal resistance a constant value?
Further, is the thermal resistance independent on the driving
current and the junction temperature?

Hope your reply, thanks a lots

For DC drive and most types of mounting, a constant thermal resistance
works well over a broad range of temperatures. If the temperature of
anything rises a lot above ambient, then an increasing proportion of
heat loss is through radiation, and the simple model loses accuracy.
Convection can also become more efficient for higher delta T.

For pulse excitation, the system is better modeled using thermal
resistances and capacitances. Some links:

http://www.ledtemp.com/Thermal%20management.pdf

http://www.ledtemp.com/Calculation.htm

http://literature.agilent.com/litweb/pdf/5091-9704E.pdf

among many such
Paul Mathews

.



Relevant Pages

  • Re: LM92 temp problems - again
    ... I have 2 lm92's connected to a controller. ... The fundamental problem with most temperature sensors is self-heating. ... 1.56C with a 200C/W thermal resistance. ... corresponds to 15mW or 3C. ...
    (sci.electronics.design)
  • Re: parallel FETs
    ... >> I can't make sense of that Terry. ... >> Shouldn't there be some thermal elasticity between FETs, ... >and will be, temperature differences between them. ... And the lower that thermal resistance between devices is, ...
    (sci.electronics.design)
  • Re: FET maximum temperature for short bursts of pulses
    ... > handled differently, junction to case, junction to ambient, and case to ... > will rise above the case temperature, but directly what the peak junction ... more closely and use the proper thermal resistance values for the formula. ...
    (sci.electronics.design)
  • Re: About thermal resistance of LEDs
    ... I have a question about thermal resistance of LEDs. ... is the thermal resistance a constant value? ... current and the junction temperature? ... or use e-mail address listed at the Web site. ...
    (sci.engr.lighting)