Settling time of FDA with resistive CMFB

[harpv]

Hi,

Firstly, I have certain basic questions which I need to be clear about before I can come to my main question :

1) Is the settling time specification given for a closed loop system?
2) Or if I have a first order system there won't be any ringing at the output right? Then we can directly see the value (say if settling time accuracy is 2%) at 98% of final value and call the time to reach this as the settling time?

So to the main question,

I have an FDA which has a folded cascode architecture (NMOS input stage) and it has a CMFB with resistive sensing.

3) Will CMFB be the major factor affecting the settling time since it forms a closed loop?
4) And since the sensing resistors will also load the FDA reducing can it also improve my settling time since it'll reduce the RC time constant??
5) Also can reducing the parasitic cap at the output also improve the settling time?

Any help is much appreciated.

Thanks & Regards,
Hari

 

[yuvan]

The settling time is VALID only for closed loop system and has no meaning for open loop system. But having said this, the UGB is defined by the open loop gain -->which defines the UGB even for the closed loop (atleast until First order approx). The CMFB cannot introduce any settling issue as long as it is properly compensated. And yes you can reduce the resistor reducingt the RC time constant, but keep in mind that you are heavily loading the main amp. So, can be reduced as along as it doesnt affect the main differential loop gain. Parasitics can again be reduced if the output is non-dominant pole (assuming it is well away from the UGB) and if so you wouldnt see a major difference in Settling time.