Designer’s Guide Community

Ankit

New Member

Offline

Posts: 1

Fully Differential Amplifier MISMATCH analysis
Apr 17^th, 2018, 2:56am

Hi all,

I am designing two stage fully differential amplifier. Both stage are CMFB compensated. My problem is that after first stage only there is so much offset in output of first stage (because of mismatch) Second stage ouputs are going to vdd and ground.

Is there any solution to it.

Please find the ckt image attached.

Waiting for reply urgently.

Thanks,
Amit

Back to top

Capture_017.PNG

« Last Edit: Apr 17^th, 2018, 6:49am by Ankit »

IP Logged

kumar.g Community Member Offline Posts: 51 Frankfurt, Germany	Re: Fully Differential Amplifier MISMATCH analysis Reply #1 - Apr 23^rd, 2018, 7:54am What is the gain of this cascaded amplifier? It could be that the gain is large enough to cause the output to rail. Since you have two dominant poles you must be careful with the stability of this amplifier.
Back to top	IP Logged

kumar.g

Community Member

Offline

Posts: 51
Frankfurt, Germany

Re: Fully Differential Amplifier MISMATCH analysis
Reply #2 - Apr 23^rd, 2018, 7:59am

Also the 2nd stage is not a differential amplifier. It acts as a common mode amplifier. Additionally, since you have alternating NMOS and PMOS input stages, check if the output common mode of 1st Op-amp is enough to keep the 2nd stage input MOSFET in saturation. This may not be relevant to your question.

Back to top

IP Logged

rajasekhar New Member Offline Posts: 7	Re: Fully Differential Amplifier MISMATCH analysis Reply #3 - May 13^th, 2018, 9:35pm Hi- WHat are you simulating here, i guess you are running open loop sim. Since it is a high gain amplifier, it is intended to use in close loop so kindly connect in Unity gain mode and check please!
Back to top	IP Logged

vroy_92

Junior Member

Offline

Posts: 30
Leuven, Belgium

Re: Fully Differential Amplifier MISMATCH analysis
Reply #4 - Jun 1^st, 2018, 3:25am

kumar.g wrote on Apr 23^rd, 2018, 7:59am:

Also the 2nd stage is not a differential amplifier. It acts as a common mode amplifier.

Agreed that it is a pseudo differential stage and will have common mode gain, but by the use of the CMFB, won't the common mode be fixed to what the input to the second stage CMFB amplifier is?

Back to top

Regards,
V Roy

IP Logged

vroy_92 Junior Member Offline Posts: 30 Leuven, Belgium	Re: Fully Differential Amplifier MISMATCH analysis Reply #5 - Jun 1^st, 2018, 3:31am Is the input pair dominating the mismatch? Then its easy to fix (blow up area of the transistor/increase the gain). But why is your second CMFB designed with the resistor cancellation network?
Back to top	Regards, V Roy IP Logged