Hello. Is anybody available now? I have a query regarding control loop stability in dc-dc buck. I am still not clear about the relation between cross-over frequency and the switching frequency and how anyone of these is related to the LC resonant frequency at the output of the Buck Converter. Can anyone please explain me. I have been reading many PDFs and articles regarding this concept and asked many questions in the forum on the same. I received good answers but unfortunately,

I am still unable to get clarity on the topic. Can someone help me?

Is anyone here?

There?

Which crossover frequency?

Gain crossover frequency. How is switching frequency of the DC-DC Buck converter related to the gain crossover frequency of the feedback control loop ?

So the basic idea is that the control loop is working at a much lower frequency than the switching frequency. That lets the control loop treat the switch-mode part as "continuous"

Now if you're talking about resonant converters, I'm not too familiar with those.

Thank you for your comment. But that's what I am not able to understand. Why should the control loop work at a lower frequency that the switching frequency of the DC-DC Converter?

Please provide some sort of simple analogy.

@Newbie looking at it from a frequency domain perspective, the input and output caps keep higher frequencies from making it into the control loop to begin with, no?

Yes.. But can you explain a little more

@Newbie think of a load transient -- if the output capacitor attenuates it acceptably, then there's nothing left for the control loop to do in response

So to start with an exaggerated example, let's wonder why we don't want the control loop frequency to be 10x greater than the switching frequency. Assuming that we fix the frequency of switching (just for a simple converter, imagine this scenario:

Our buck switch has turned on, and turned off, and we are in a continuous mode where current is flowing through the inductor and the diode in a typical buck converter. Suddenly (and magically) we detect that our output voltage has dropped, and needs to be compensated. If your only "knob" to turn is PWM duty cycle, you've got just about a full switching period before you can begin to address the error signal.

Now, to address that "magical" part, the output voltage will have ripple at the switching frequency. This is normal and expected. A simple controller cannot distinguish between the peaks and troughs of the normal ripple and steady-state error. If the circuit is permitted to use this ripple in PWM compensation, a similar issue to our "nonsense" scenario occurs.

What we really want to be doing is taking the average output voltage, and comparing that to our reference. Because we want the average, then we really don't want a known source of noise and error (the output voltage ripple) to significantly impact our PWM duty cycle adjustment.