PWM current impact on power supplies

[Rigolon]

Hello everyone,

I'm having some doubts on some details of how a PWM signal works. I understand some of the basic knowledge, e.g. a component that is capable to withstand a maximum forward current of 100mA is able to withstand 1A if the signal has a small duty cycle.

What I am confused is about:
* This is applied for every component?
* If the maximum current rate is 100mA, I could use any current as long the Irms = 100mA? For example use a 1A with 10% D.C. ou 2A with 5% D.C.?
* Is there a frequency limit where this 10% D.C. of 1A draws more than 100mA, or it doesn't matter the frequency it will always draw 100mA? (In this supposition the components immediately rise and fall with the signal, it does not have rising or falling time) 
* And the most important thing, if I use a 10% D.C. signal at 1A, do I have to use a 1A power supply or does a 100mA could be use?

[oPossum]

In switching power supplies using MOSFETs the current will typically be well below the continuous rating to ensure it stays out of the saturation region.

Infrared LEDs can usually handle substantial pulse currents. Up to 1A for a 5mm package is not unusual.

You have to carefully read the spec sheet for a specific part to know what it can do.

[P_Doped]

Components which are often subjected to pulsed currents will have a separate pulsed current rating.

This current is not as simple as your PWM fractional scaling, unfortunately.

The limiting factors can include internal element heating, electromigration, and other physical phenomena.

But, the good news is that the pulsed current will (should?) be higher.

As Raven said, check component datasheets.

I did 1 quick check on an n-channel MOSFET and it was rated for 10A continuous and 26A pulsed (pulse width limited by thermal temperature ratings)...