The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: Victor Roberts
Organization: Roberts Research & Consulting, Inc.
X-Mailer: Mozilla 4.78 [en] (Win98; U)
Subject: Re: luxeon star leds
References: <3DDD9A57.58900EFB@RobertsResearchInc.com> <email@example.com> <3DDDB646.96D0D9D2@RobertsResearchInc.com> <3DDE7264.CD48F4F8@RobertsResearchInc.com> <firstname.lastname@example.org>
Date: Fri, 22 Nov 2002 22:31:39 GMT
NNTP-Posting-Date: Fri, 22 Nov 2002 14:31:39 PST
John Larkin wrote:
> On Fri, 22 Nov 2002 18:07:32 GMT, Victor Roberts
> > Operating [[without]] a heat sink will decrease the
> > apparent thermal resistance between the junction and the
> > heat sink mounting plate on the LED because of the other
> > paths available for power loss. Operating without a heat
> > sink does not lower the junction temperature, it raises it.
> > No, I mean what I said in my last message. Operating [[with]] a
> > heat sink _decreases_ the "apparent" thermal resistance
> > between the junction of the LED (or any other semiconductor
> > device) and the heat sink mounting plate on the LED (or
> > other semiconductor) package.
I apologize for the fact that this comment of mine has lead
to so much confusion. Perhaps an example will help to
explain what I am talking about.
I am going to start with data from the Luxeon Star data
sheet. Here are some basic numbers for the white Luxeon
Forward current = 0.350 Amps.
Forward Voltage (Typical) = 3.42 volts.
Thermal Resistance (Junction to board) 17 deg C/W. The
"board" is the aluminum plate of the bottom of the Luxeon
package. It is 25mm on a side.
Output (Typical) = 18 lumens.
From the values given by Lumileds we can calculate the
Power (Typical) = 1.20 watts.
Efficacy (Typical) = 15.04 lm/W.
Based on this efficacy, the LED converts electrical power to
light with an efficiency of 5% to 10%. I will use 8% for the
rest of the calculations, but the exact number is not
important. I will also assume that Lumileds followed
standard practice and used the power dissipated by the chip,
not the power delivered to the chip, to calculate the
thermal resistance. However, since the conversion efficiency
of the LED is low, they could have used the delivered power
and the results would have been off by only 5% to 10%.
If we further assume that the only radiation emitted by the
LED is the visible light, then we have an input power of 1.2
watts (as calculated above), 0.1 watts of white light (0.08
* 1.2) and therefore 1.1 watts of "waste" power.
There is more than one dissipation mechanism for the waste
power. When the Luxeon Star is mounted on a heat sink at 25C
the great majority of the waste power will be conducted to
the "board" and from there to the heat sink. However, a
small fraction of the power will be dissipated via
convection from the top of the LED package. The thermal
resistance combines all the power loss mechanisms into one
number and is "presented" as if all the waste power flows to
the heat sink. If the dissipated power is 1.1 watts and the
junction is 18.7 C hotter than the "board" then we can
calculate that the thermal resistance between the junction
and the "board" is 17 C/W. If the "board" is 25C, then the
junction temperature will be 25 + 18.7 or 43.7C.
Now, if we take the Luxeon off the heat sink the temperature
of the whole device, junction and heat sink will go up.
Assume we continue to pump 1.2 watts into the LED and that
the conversion efficiency of electricity onto light does not
change (which is not correct because the Luxeon data sheet
shows that the conversion efficiency drops as the junction
temperature goes up). When the whole device is at higher
temperature than it was while mounted on the heat sink, then
a greater percentage of the 1.1 watts will be dissipated
from the top side of the package via convection, leaving a
slightly smaller amount of power flowing to the "base" via
conduction. Since the actual thermal resistance between the
junction and the "base" does not change, the reduced power
flow between junction and base will reduce the temperature
differential between the junction and the "base". Therefore,
the thermal resistance between junction and base will appear
to decrease, since for the same power we now have a lower
So many words for such a small effect.
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup