Simulating the slew rate with capacitive feedback fully differential amplifier

[Junus2012]

Hi,


Why reverse?

The bigger the C the lower the impedance.

Usually: A = R_fb/ R_in....in your case it is: A= X_fb / X_in.....nothing reversed.

Klaus

Thanks klaus, you are right, I was confused cause in my simulation with out a resistor it was C1_fb/C1_in, but now as you explained it becomes impedance relationship

 

[sutapanaki]

It is missing the symmetry probably because you didn't simulate long enough like in your post #14. Can you run longer simulation?

If the intended application is to build a sample and hold circuit, than you are going on a wrong path. Sample and hold circuits using OTAs are a switched capacitor circuits and then just by switching the caps in 2 phases usually, you define also the OTA input common mode. I think you should build a test bench that is closer to the application you have in mind. This cap-cap OTA you have right now is good for running .ac analysis for the S/H but not for transient simulations.

 

[Junus2012]

It is missing the symmetry probably because you didn't simulate long enough like in your post #14. Can you run longer simulation?

If the intended application is to build a sample and hold circuit, than you are going on a wrong path. Sample and hold circuits using OTAs are a switched capacitor circuits and then just by switching the caps in 2 phases usually, you define also the OTA input common mode. I think you should build a test bench that is closer to the application you have in mind. This cap-cap OTA you have right now is good for running .ac analysis for the S/H but not for transient simulations.

Dear Suta,

I have simulated with longer time, as you expected the signals are becoming true after a while and looking fine as shown below



Now this issue is being cleared,
Now I am interested in your suggestion to tun the simulation with two phases, I never tried before this kind of simulation used usually for SC circuit, could you please help to setup this test and how it looks

Thank you

 

[sutapanaki]

Your amplifier right now is not suited for that. If you want to do it, you have to use the OTA in a SC configuration. As it is now with that resistor in the feedback is not quite good for S/H because you've probably noticed in your simulations that the pulses at the output droop with time because they lose charge through the capacitor.
You could look into a flip-around S/H which is one of the not so complicated configurations. SC S/H circuits are also suitable for use with SC CMFB.

 

[Junus2012]

Your amplifier right now is not suited for that. If you want to do it, you have to use the OTA in a SC configuration. As it is now with that resistor in the feedback is not quite good for S/H because you've probably noticed in your simulations that the pulses at the output droop with time because they lose charge through the capacitor.
You could look into a flip-around S/H which is one of the not so complicated configurations. SC S/H circuits are also suitable for use with SC CMFB.

Dear Suta,
Thank you for your help

I have two questions,

1. I have designed my OTA with continuous time CMFB circuit, will it be a problem or it is only optional to design SC CMFB ?
2. Are switches available by Cadence ?

Thank you

 

[sutapanaki]

It can probably work with CT CMFB. Check it in simulation.

What do you mean by "switches" in Cadence? Switches are transistors and transistors are available in the design kit.

 

[Junus2012]

It can probably work with CT CMFB. Check it in simulation.

What do you mean by "switches" in Cadence? Switches are transistors and transistors are available in the design kit.

Thsnk you Suta,

I mean ideal switches

 

[sutapanaki]

Yes, it is called "switch" I think in analogLib.