TPS55289: crossover frequency

lelian cao

Part Number: TPS55289

hi team:

We recently tested the frequency response of TPS552892QWRYQRQ1 circuit attached, the results shows phase margin (61.7°) and gain margin(13.6db) pass the criteria, but crossover frequency is lower than 10kHz, which means the bandwidth of this circuit is lower than 10kHz. And we found most of buck-boost has low crossover frequency.

My question is:

1. Is low bandwidth (lower than 10kHz) of regulator acceptable if load transient meets requirement?

2. Do buck-boost and boost have lower bandwidth than buck in general? if yes, why?

3. I can adjust crossover frequency by optimizing compensation network, but the optimized bandwidth should be limitied by the regulator itself no matter how I change the comp network, so how do I know the bandwidth limitation of the regulator?

Thanks.

、

over 1 year ago

0 Mulin Yuan over 1 year ago

TI__Genius 15160 points

Hi Lelian,

Thank you for reaching out.

1. It is ok if the load transient meet requirement.

2. In boost mode, there is a right half plane zero, which has negative influence in PM/GM. Because of the right half plane zero, we always set cross frequency relatively low to avoid the influence. So generally boost has lower cross frequency.

3.To increase cross frequency, you can increase Rcomp. But this may cause PM/GM decrease. As far as PM/GM/load transient meet requirement, the compensation parameter is ok. No need to pursue high cross frequency.

Regards,

Mulin

0 lelian cao over 1 year ago in reply to Mulin Yuan

Prodigy 160 points

Hi Mulin:

Thanks for your answer.

I have one more related question, we found buck has lower crossover frequency in DCM mode with light load, I would like to know what causes the reduction in bandwidth in DCM, and how it influence the bandwidth theoretically.

Thanks.

0 Bryce Xun over 1 year ago in reply to lelian cao

TI__Mastermind 27830 points

Hi Lelian,

It will take us some time to check the Buck DCM transfer function, or you can post a new thread including Buck team to help answer this.

BRs,

Bryce

0 lelian cao over 1 year ago in reply to Bryce Xun

Prodigy 160 points

Hi Bryce:

Thanks for your time, I will wait your reply in this thread.

This is what I found from related slides of TI, I wanna know more of the details about the conclusion, like why is this inductor current behavior reduce bandwidth in DCM.

Thanks.