# MOSFET temperature curve?

Discussion in 'General Electronics Chat' started by nepdeep, Nov 2, 2012.

1. ### nepdeep Thread Starter Active Member

Sep 14, 2011
140
0
Hi everyone,
I need some help understanding this curve....
This is the thermal characteristic of mosget...

2. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
510
It;s actually a plot of thermal IMPEDANCE versus pulse width (in time duration), which is to say pulse width of the power pulse which is creating the heat. Thermal impedance, or thermal resistance, is how much resistance there is in heat flowing from the die to the heat sink. It varies with pulse width because a very narrow puls just heats the local mass but as the width increases, it reaches steady state heat saturation of the mass and it is continuous flow of heat. For that reason, the aparrent thermal resistance is much lower to a very narrow power pulse than it is for a wider pulse or continuous pulse.

For this device, steady state saturation is reached when the pulse width gets longer than about 100 ms.

3. ### Ron H AAC Fanatic!

Apr 14, 2005
7,049
675
We should add that the numbers inside the plot, on the left side (single pulse, 0.01, 0.02, etc.) are duty cycle (1%, 2%, etc.).

4. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
510
yep, duty cycle of the pulse train.

5. ### nepdeep Thread Starter Active Member

Sep 14, 2011
140
0
Okey..but at what power is the curve valid...cause earlier i used the thermal resistance to calculate the maximum power...now it seems like i calculate duty cycle ...is that at the maximum current and voltage...or under possible maximum power dissipation based on the thermal resistance and maximum junction temperature of the mosfet?
hope to hear form you soon...

Last edited: Nov 2, 2012
6. ### nepdeep Thread Starter Active Member

Sep 14, 2011
140
0
Is tp the total time period on the graph...
At the moment what I think is...
1. Calculate the Pmax based on the thermal characteristic of mosfet.
2. Now choose certain duty cycle...
3. again choose the thermal resistance and
4. again determine the maximum power dissipation..

M little confused...we get different values from kind of same calculation..i really need some help...thanks for all the help

7. ### bountyhunter Well-Known Member

Sep 7, 2009
2,498
510
Not sure how to explain. A good way to look at it is the thermal resistance is a specific value (on the data sheet) like for a T0-220 it would be typically about 3C/Watt for CONTINUOUS POWER.

Transient thermal resistance is lower for pulses, but only for pulsed power.

For a safe design, just use the Rth value for continuous power. Pulse ratings only are useful in apps like a switching converter where switching losses generate very fast and narrow power pulses at turn on and turn off intervals. They have no use for any kind of continuous power application.

It would help if you actually described what you are trying to do and why this graph came up.

Sep 7, 2009
2,498
510