Phase margin of 2 stage cmfb diff amp and its compensation

[Safiya]

I have designed 5MHz bw 2stage CMFB diff amp. but i could not adjust the Phase Margin using R,C values for the compensation,

in my case when i achive PM = 60 , bw reduces to 800kHz. is there any method to compensate this amp of BW 5M or 3MHz with enough PM.

Regards,
Safiya

 

[FvM]

You'll primarly look at the uncompensated loop gain to decide what's feasible.

 

[Safiya]

You'll primarly look at the uncompensated loop gain to decide what's feasible.

yes , I have checked with uncompensated gain and bw. which was OK (50dB),

but with phase margin it sticks to -58 or 296 deg , and BW reduces when i try to optimize with lower cap. of miller compensation. R, C

 

[tompham]

Hi Safiya
You use folded cascode design to save power and easier to compensate.

 

[Safiya]

initial uncompensated Phase is -179deg, for whole freq band it is negative as well around ,
what is the reason behind and what value we should use initially to compensate for , R, C without affecting BW.

 

[nitishn5]

Compensation would always reduce the bandwidth since it makes one pole more dominant and moves the other pole(s) more non-dominant. So bandwidth would always be less.

If you want to have the same bandwidth as the uncompensated one, you would have to increase the current.

For a simple Miller Compensated Amp, UGB is gm/Cc. So you would need to increase the tail current to increase gm which would increase your bandwidth.

You cannot reduce Cc since it is what achieves the pole splitting and gives you the Phase Margin.


Of course there would be other techniques where you can not burn a lot of extra current and still have a stable loop, but this one is the simplest and hassle free.