Transient thermal Z for IPD90N04S4L-04

[Techman_7]

For Si MOSFET IPD90N04S4L-04.xls (infineon.com), the transient thermal impedance graph is as below -

Are all these curves for single pulse? Or repeated pulses?

I want to calculate the peak pulse current & power dissipation capacity for this MOSFET for specific pulse duration for repeated pulses.
If above graph is for single pulses, then for repeated pulses how to calculate power dissipation?

 

[dick_freebird]

It appears to me that only the lowest curve is (as it says)
single pulse. The others are labeled with what I think is a duty
factor.

So take whatever your worst case on time (plus probably the
Miller plateau intervals' heat-slug on either side) duty factor is,
pick your line, use the pulse time axis, bickety-bam.

You want to understand their test config and conditions as the
gate drive can be a big factor in switching losses.

 

[KlausST]

Hi,

I admit it could be clearer.
It says parameter D (duty cycle) = tp (pulse length in time) / T (period time)

Duty cycle is only meaningful on repeated pulses.
So the bottom line is for single pulse, all others are repeated pulses.

The pulse frequency need to be high enough so the thermal capacity is able to keep thermal ripple low. f = D / tp

Klaus

 

[D.A.(Tony)Stewart]

Note that Zth reduces in half for d= 0.5 or 50% duty cycle.
Then again Zth is 10% of continuous for d=0.1 or 10%
So steady-state duty-cycle has a linear effect on steady-state Zth.

Note that there are asymptotic time constants that must correlate with each layer of material.

The straight-line asymptotes appear to me as 1 ms, 10 ms and 100 ms, so transients in your d.c should consider the worst-case and initial conditions.