Resistive Feedback Bode Plot Problems

[Arielle]

Hi guys,
I have a very fundamental problem: First, is it possible to use resistive feedback on a circuit as shown in the picture? When I tried to simulate it, the bode plot had a very odd form and the DC gain I calculated by -Rf/Rin could not be seen in the simulation results... Help pls.

Arielle

 

[crutschow]

Those are not identical circuits. The circuit on the right is a simple differential to single-ended amplifier which has a DC offset between the input and output. You can not use straight resistive feedback with that. You need a level shifter output stage and frequency compensation to operate as an op amp, similar to that shown in the left picture.

 

[Dominik Przyborowski]

Your OTA is not bufffered, so by adding resistive feedback You decrease an open loop gain of amplifier from gm2(rds2||rds4) down to gm2(rds2||rds4||Rf). So if Rf<<(rds2||rds4) a small signal gain of amplifier goes below 0dB. In addition if You have unipolar power supply (e.g. gnd and +vdd only) You should put your small signal ground on some level, e.g. vdd/2 to provide correct dc operating point. Of course You should remember about output current limitation - your feedback current shouldn't be higher than current flowing in amplifier branches.

 

[Arielle]

Hey, thank you two! I tried to do as you said (see screenshot) and I think that my gain can be set by Rfb/Rin now. However, my phase is... not what I expected and if I apply larger AC signals than 1mV at the input, my output common mode goes up. I made pictures and hope, you can tell me what I did wrong... ;__;

 

[Dominik Przyborowski]

Now You added second amplifying stage so you have to add compensation RC network to ensure stability (look for miller opamp). I mean buffered stage → simple source follower to provide low output impedance to drive resistive feedback.

 

[LvW]

Hey, thank you two! I tried to do as you said (see screenshot) and I think that my gain can be set by Rfb/Rin now. However, my phase is... not what I expected and if I apply larger AC signals than 1mV at the input, my output common mode goes up. I made pictures and hope, you can tell me what I did wrong... ;__;

Arielle, I really don`t know what you are asking.
I see a differential amplifier (with a high resistive output: OTA) and a second stage.
Please note that one amplifier input node is OPEN !
You speak about feedback, but where is your feedback path?
What is your goal? Amplification with resistive negative feedback ?
Why do you inject a signal into the OUTPUT?

 

[Arielle]

Hi guys, just wanted to say that it's finally working & thank you very much for your help. After I inserted a buffer (instead of a second stage hahaha), it worked out =)
@LvW: sry, my screenshot was bad because it didn't show that I had actually a signal at the second input, so it's not open

 

[LvW]

Hi guys, just wanted to say that it's finally working & thank you very much for your help. After I inserted a buffer (instead of a second stage hahaha), it worked out =)
@LvW: sry, my screenshot was bad because it didn't show that I had actually a signal at the second input, so it's not open

OK fine. So it is "working". And you are sure that the results are correct?