What is smallest period for PWM?

So, I’ve been playing around with PWM, and I would like to have one PWM channel drive a buck circuit and another channel drive a boost circuit.

The magnetics that I am looking to use would like to be driven in the 100-150khz range. Mosfets to drive this are not a dime a dozen, but perhaps $10/dozen.

I’ve read other places that the PWMs can have a base frequency as high as 1.6Mhz?

Currently I’m using P9_14 to drive the input of an oscilloscope which is 1Mohm-15pf and the wave form looks like slow capacitor charging. When I model the slope on spice it looks like the output impedance of P9_14 might be in the 50-70kOhm range?

This means that I can’t create a crisp pwm signal at any where near the frequency I’m hoping for.

Also, this would mean that the maximum current out of a pin at 3.3v would be 0.06mA… and that has got to be wrong as I have seen too many warnings to limit source/sink current to 6mA or less???

Any ideas what I might be doing wrong that is limiting my pwm max frequency so badly?



EHRPWM functions of AM3358 are running with 100 MHz. Thus the smallest possible half pulse width is 10 nsec (10 nsec of and 10 nsec on for a full period).

Thanks Karl.

Right on. So a 50Mhz frequency with a 10ns pulse should be possible! Great!

Have you been able to observe such a waveform on your BBB?

The issue that I am having is that I can’t get the pin on the Beaglebone Black to move nearly that fast!

Right now, a 10,000ns pulse hardly even registers as a blip on my oscilloscope… so what I am seeing is 3 or 4 orders of magnitude off from what you suggest I should be able to see.