Loop stability understanding of buck regulator

Greetings people.

1) I need a suggestion for a good book that focuses mainly on buck regulator design, single and multi-phase and with good explanation of type-ii and type-iii compensation.

2) I am trying to understand a simple buck regulator's loop stability with a type-ii or type-iii compensator, not in terms of equations, but a more intuitive understanding.
The LC filter creates a 'double pole'...what does it mean in terms of time-domain?
And how will adding this 'compensator' help avoid the instability?

vrs007 said:

Greetings people.

1) I need a suggestion for a good book that focuses mainly on buck regulator design, single and multi-phase and with good explanation of type-ii and type-iii compensation.

Click to expand...

A book is probably overkill, since compensation is relatively simple. Here is an excellent white paper on compensation of voltage mode supplies:
**broken link removed**

Keep in mind that a type III compensator is only really necessary in voltage mode, and won't help much for current mode.

2) I am trying to understand a simple buck regulator's loop stability with a type-ii or type-iii compensator, not in terms of equations, but a more intuitive understanding.
The LC filter creates a 'double pole'...what does it mean in terms of time-domain?
And how will adding this 'compensator' help avoid the instability?

Click to expand...

It's far far easier to analyze things in the frequency domain than time domain, so stick with that.
Poles cause gain to decrease at 20dB/decade and will (usually) cause a 90 degree phase lag. Zeros do the opposite, causing gain to increase at 20dB/decade and (usually) phase will lead by 90.

When looking at stability, you are mainly interested in avoiding the condition where loop response lags by -360 degrees (or -180 degrees if you leave out the inversion of the inverting error amplifier) while loop gain is greater than 0dB. When that condition occurs, you will get positive feedback and oscillation. So if your supply has two poles, then its phase lag will quickly drop to -180 above that resonant frequency and will approach instability.