From: firstname.lastname@example.org (Bob Wilson)
Subject: Re: Question about linear derating factor for MOSFETs
Date: Mon, 11 Nov 2002 04:36:43 -0000
Organization: Your Organization
X-Newsreader: WinVN 0.99.9 (Released Version) (x86 32bit)
In article <email@example.com>,
>The IR datasheet for a MOSFET I was considering (for a larger version
>of a constant-current circuit that this group is giving me great tips
>on) has the Linear Derating Factor listed as 5.0W/degree C. The max
>power dissipation rating is 625W (SOT-227 case) at Tc = 25 degrees C.
>I can't find what "Tc" (subscripted "c") means anywhere on the
>datasheet...is it case temperature? For device derating purposes, can
>I assume that the heat sink temperature (using the delta-Temp
>calculated using the load and sum of the thermal resistances from the
>junction to air) will be the same as the case temperature?
>If that's true, I think the derating is applied this way:
>If the heat sink temp is 80 degrees C. under load, then I have a
>55-degree difference (from 25-degrees). Taking the 5.0W/degree C.
>derating factor gives me a 275W derating. I would subtract that from
>the 25-degree rating of 625W to arrive at the 80-degree rating of 350W
>for the device.
>This means, for this device, that I cannot safely have a load greater
>than 350W with the heat sink I'm using? Assume that I've maximized
>the effectiveness of the other factors involved...fan, possible
>insulator (grease, Kapton, Sil-Pad, etc.), etc.
Tc is standard industry nomenclature for care temperature (it has been
universally used for 30 or more years.)
The heat sink temperature will most certainly not be the same as the case
temperature, unless there is no heat flow. In your case, the difference can
be very sunstantial. A typical thermal resistance for a an "Isostrate"
insulated case-to-heatsink interface will be around 0.2C/Watt. The other
extreme is a Silpad which is (at this power level) basically useless, and is
several degrees C/Watt.
So let's take the 0.5C/W as an example. Pumping 350W across the
case-to-heatsink junction will result in a 175 C (!!!!) temperature drop
across this interface. In other words, your case temperature will be
175C above the heatsink temperature! This clearly tells you that there isn't
a hope of approaching the power level you are mentioning, using only one