I am trying to calculate the maximum single peak pulse power rating of a thick film resistor that will be subjected to a significant amount of inrush current upon each power up.

The datasheet has a nice plot of maximum power vs surge duration for each package size on page 2, but I am struggling to determine the appropriate way to determine the surge duration of my transient.

The transient has somewhat of a double exponential shape, what interval would typically be considered the pulse/surge duration of a waveform like this?

My particular transient has the following timing characteristics:

  • t_1 = ~5 ms (rise time to peak value, 0%-100%)
  • t_2 = ~15 ms (cumulitive time to reach the 50% decay point)
  • t_3 = ~95 ms (total time until transient is effectively 0 again)

I feel like I remember learning that you measure out to the point at which the amplitude has decreased back down to 50% of the maximum, so in my case, 15 ms, but I can't find any sources online that back that up.


1 Answer 1


So I believe I have found the answer from some of my colleagues, they stated that the typical approach is to measure the time from the beginning of the pulse up to the point at which it has decayed back down to 50% of the peak value.

Alternatively, I came up with an alternative solution as well. Since the total power of the waveform is equivalent to the area under the curve over its entire duration, if one were to take this value and calculate the duration of a purely rectangular pulse that had the same peak value and the same total power, this pulse length would be a safe estimate as the damage potential of the rectangular pulse would be greater than a longer exponentially decaying one as the same total power is contained in a shorter duration (at least for the case of a purely resistive load with no complex impedance term).

  • 2
    \$\begingroup\$ "Since the total power of the waveform is equivalent to the area under the curve over its entire duration ...". Nope. That's the energy. (e.g. kWh or joules.) The rest is probably OK. \$\endgroup\$
    – Transistor
    Sep 7, 2017 at 21:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.