Stability determination from step response

M@rtin

Hello everybody,

in this paper(http://focus.ti.com/download/trng/docs/seminar/Topic%207%20-%20Evaluation%20and%20Performance%20Optimization%20of%20Fully%20Integrated%20DC-DC%20Converter.pdf) I found on page 7-2 I can use the step response to determine the phase margin in the vicinity of the crossover frequency. Is it not enough to determine whether the system is stable? I think I dont need to know the gain and phase for lower frequencies or I do?

over 12 years ago

0 Chris Glaser over 12 years ago

TI__Guru*** 147960 points

You'll need to do whatever you need to do to make you and your company feel comfortable with the design performance. There is no minimum set of data required for every situation.

Here are two more app notes that are relevant to this topic:

http://www.ti.com/lit/an/slva289/slva289.pdf

http://www.ti.com.cn/cn/lit/an/slva381a/slva381a.pdf

0 JohnTucker over 12 years ago in reply to Chris Glaser

TI__Guru**** 191445 points

The step response is the best criteria for stability. Form that you can tell if your system is over, under or critically damped. Personally, i use both the loop response (Bode) plot and transient analysis when checking out my designs.

0 M@rtin over 12 years ago in reply to JohnTucker

Prodigy 110 points

Thanks for reply and relevant aplication note. So I can use only the step response to determine the system stability. For me is easier measure the step response than the bode plot. I design power supply only once a year and I dont need to feel more confortable.

Power management

Power management forum

Stability determination from step response