Subject: Re: Question about linear derating factor for MOSFETs
Date: Tue, 12 Nov 2002 21:43:00 +0000 (UTC)
NNTP-Posting-Date: Tue, 12 Nov 2002 21:43:00 +0000 (UTC)
User-Agent: tin/1.5.14-20020926 ("Soil") (UNIX) (Linux/2.4.19 (i686))
In sci.electronics.design John Larkin wrote:
> yeah, the actual junction thermal time constants are pretty short. But
> there is a (complex) composite time constant that is the junction, the
> small heat spreader in the package, and the copper (or whatever) in
> the heat sink itself. The heatsink situation is one of those nasty 3D
> thermal diffusion things that is hard (at least for me) to begin to
> quantify. In the 150 msec sort of domain, I'm thinking the thermal
> dynamics of the heatsink matter.
> I've often wondered what might be...
> The thermal resistance and time constant of a TO-220
> bolted to an infinite sheet of 0.062 aluminum
> Ditto, a TO-247 bolted to half-a-universe full of copper
> And a few similar situations that would be handy starting points.
> Too bad the serious thermal simulation software is so expensive.
You can ballpark really simple things like this.
Now, look up the thermal properties of the element in question.
For example, copper has a thermal resistance of 400W/m/K, a density of
about 9000Kg/m^3, heat capacity of about 385J/Kg/K.
Now, work out the heat capacity per unit volume:
Recasting this in terms of cm, as not many semiconductors are meters across...
Heat capacity 3.6J/cm^3/K
Thermal resistance 4W/cm-1/K
Let's look at the first millimeter next to a centimeter junction.
Assuming it's heated by 100K, it'll take about .36J*100 = 36J to heat it up.
At the start of the period, it's heated by 100K on one side, leading to a heat
flow of 4Kw, if this heat load were sustained.
As the heat flow would have heated the first millimeter by 100C in about 10ms,
it's probably fair to say that initially at least, the heat propagates at
If you think carefully, and apply basic geometry, you can get usefull answers.
An infinite heat-sink is not going to be able to conduct heat better than
the first centimeter.
http://inquisitor.i.am/ | mailto:firstname.lastname@example.org | Ian Stirling.
Paranoia: A game for the whole family, and anyone else who might be watching.