# Meaning of “pulse duration” in transistor datasheets?

I have a somewhat basic question. Many transistor datasheets have something like this for the peak current:

pulse duration = 300 μs, duty cycle ≤ 1 %

What does pulse duration mean? Is it only the on time, or is it the time period (on + off time)?

so, for above, which is correct?

A) 3us on, 297us off

or

B) 300us on, 29.7ms off

I'm guessing B is more useful, but A might look better on paper (except if transistor is too slow). So I'm somewhat confused (not having any engineering background and all...)

Huh, it never occurred to me that anyone would interpret it as (a).

I believe it is (b). 300us tells you the on-time, 1% implies the period (how frequent you can do it).

This is because it says pulse DURATION, and a pulse exists only while it is on.

• Thanks. That makes sense. There also appears to be some more terms floating around... pulse period, inter pulse period... I guess pulse duration is same as pulse width, and pulse period same as time period of the "pulse train". – Indraneel Nov 11 '19 at 13:17

It means that it's on for $$\300\mu \mathrm{s}\$$, because anything more would fry the transistor. Then off for a Good Long While to let the transistor cool down.

[example of Finite Element Modeling of thermal flows, at end of answer, for 10 micron mesh]

The thermal timeconstant of 1 micron cube of silicon is 11.4 nanoseconds.

The thermal timeconstant of 10 micron cube of silicon is 100X slower, at 1.14 microSeconds.

The thermal timeconstant of 100 micron cube is another 100X slower, at 114 microseconds.

Thus a 300uS pulse should be adequate to cause substantial heating, hopefully not to failure, of power silicon devices that are back-grinded to be 100 micron thick.

By the way, the thermal timeconstant of a cubic meter of silicon is 11,400 seconds.

simulate this circuit – Schematic created using CircuitLab

• It looks interesting, but unfortunately, I don't understand much of what is going on here. Is it that thermal resistance increases as square of distance (same as what happens with light)? (that is what it looks like since your picture does not show the heating element actually embedded inside the silicon. I'm guessing that some kind of blotzmann law would take over if it was actually embedded.... but I'm actually a biology guy, so don't have any clue). – Indraneel Nov 12 '19 at 14:13
• I edited the schematic, adding an INTERNAL HEAT SOURCE. – analogsystemsrf Nov 13 '19 at 5:20