Frequency response of chopper stabilized amplifier

[Henry98]

Hello everyone,
I am newbie for analog circuit design, I'm studying the technicque to reduce 1/f noise (flicker noise) in amplifier. which is chopper stabilization technicque.
1. I want to analysis the gain, PM, stability, noise of chopper stabilized amplifier, How I can do that in cadence tool?
Below is my circuit:

2. when I tried to AC analysis in cadence with gain and PM, The tool given me a kind of strange result look like below. Could you help me explain what is my error in simulation?

Thank you,

 

[FvM]

To determine frequency response of modulated circuit, you need to use PAC analysis rather than simple AC. Did you?

 

[Henry98]

To determine frequency response of modulated circuit, you need to use PAC analysis rather than simple AC. Did you?

Hello,
Thank you your answer,
I didn't run PAC analysis.
I just run simple AC analysis like as a normal amplifier and then use calculator with function of dB20(VF(OUT)/VF(IN)) to find gain and PM.
what is the PAC analysis? when I need PAC analysis?

Thank you

 

[FvM]

You need PAC and related analysis modes for circuits that involve frequency conversion, e.g. mixers or chopper amplifiers. There special analysis modes to simulate noise of modulated circuits. Consult Cadence Spectre user manuals and tutorials about it.

I wonder how the chopper circuit is modelled in your simulation? Also what's the simulation setup to assure correct amplifier bias?

 

[Henry98]

Your answer was helpful to me
My chopper circuit was referred from the paper, but there is only explanation that the chopper circuit at input signal is modulator and two other is demodulator.
Theoretically, chopper amplifier includes two chopper circuit : one for input and other one for output.

This picture that I cited from paper "A CMOS Chopper Amplifier" CHRISTIAN C. ENZ, STUDENT MEMBER, IEEE, ERIC A. VITTOZ, MEMBER, IEEE, ANDFRAN~OIS KRUMMENACHER.
I also isn't clear about that and I just studying about that structure.
Thank you.

 

[dick_freebird]

As a practical matter, chop-amps have an output filter
to tamp down the HF spurs as much as possible. That
will greatly alter (lower) the bandwidth of the final
assembly.

Are you only concerned about the core amplifier's
AC response or are you after the "whole enchilada"?

 

[Henry98]

As a practical matter, chop-amps have an output filter
to tamp down the HF spurs as much as possible. That
will greatly alter (lower) the bandwidth of the final
assembly.

Are you only concerned about the core amplifier's
AC response or are you after the "whole enchilada"?

Hello dick_freebird,
I’m also interested in the LPF placed at the output. I plan to add an simple amplifier at the output as shown in the diagram below. By connecting the capacitor in this way, I can create a low-pass filter (LPF) without needing a large capacitance (Miller compensation effect).

 

[FvM]

Diagram misses feedback path. Do you plan overall feedback including output filter? It can be stable if the A1*A2 has a dominant pole.

 

[Henry98]

Diagram misses feedback path. Do you plan overall feedback including output filter? It can be stable if the A1*A2 has a dominant pole.

Currently, I haven't a plan for overall feedback. I'm weak in knowledge of this part of stability. Thus, Could you please explain your idea?

As my point, A1 is a amplifier ( fully differential folded cascode) and itself also contribute a dominant pole for overall system.
And A2 and capacitor around it which is simply using miller effect to create a LPF and also eliminate the zero pole in system.
Thank you