Strange Bode Plot form a constant on time DC-DC Converter

[jwfan]

Hello,

We got a strange bode plot for a constant on time DC-DC Buck converter. This converter shows the phase crosses the zero line at low frequency. This attached waveform is the simulation results of Simplis. the network analyzer shown the similar result. Both transient simulation and test results shows stable output ...

We are confused. Could you help us? Thanks a lot,


Green curve is the Gain. Blue curve is the phase

 

[FvM]

The bode plot is involving a number of questions.

- You are apparently simulating a switching regulator in frequency domain. How? Does the simulator generate a continuous-time equivalent circuit? Otherwise it's just erroneous.
- I presume, the diagram is meant to show the loop gain. How do you measure it?
- What's the converter's switching frequency?
- At first sight, there's no impact of an output filter in the bode plot. How is it possible?

Needless to say, that the right half of the bode plot is meaningless.

 

[mtwieg]

What exactly is the system under test? Is it current mode? Loop response, plant response, or control response?

The gain plot looks similar to what you'd expect for a type III compensator. That phase response looks totally nonsensical though. At the very least it looks like there's a sign error...

 

[Alan Huang]

Hello,

I also got the same frequency response of COT buck converter. The phase looks weird.

I guess because there is no ideal trigger signal for POP in COT control. Not like current mode buck, there is an ideal PWM set signal in each switching cycle.

Does anyone can help us solve this problem? Thanks.