op amp simulation help

[simon0123]

I have draw a folded cascode opamp and I want to simulate it open loop gain and phase.

The way that I used to simulate an opamp is to connect a AC source to the non-inverting input and connect a 1F capacitor to inverting input. Also connect a 1G resistor between output and inverting input. (P.S for the AC source, I just use a VDC source and set DC voltage =0V and AC magnitude = 1V)

However, I can't get the expected results. Please help me to point out the problems of my circuit. The gain is very low and the waveform is not correct. Thanks a lot
(P.S to way that i plot the result is just click Results -> Direct plots -> AC magnitude and phase.)

Is it my simulation setting is wrong??

 

[frankaurora]

you should set CM voltage, ie. vdc>0

 

[checkmate]

Try using a 1GF capacitor. Even better, replace the 1GOhm resistor with a 1GH inductor.
Before you run AC, run a DC to ensure all the currents are as expected and all transistors are in the correct operating point.
And as frank has mentioned, you need a common-mode input which should be dependent on the application of the opamp. If you do not have an application, just use half vdd.

 

[simon0123]

i make the changes, and the follwoing is the simulation result. I think this is the correct result, right?

Actually, i want to change the single end output to differential outputs which will be used as integrator in sigma delta ADC. So, I want to know how to simulate an fully differenetial amplifier (gain and phase first)????

 

[checkmate]

You already know how to break a feedback loop via a large LC filter. The idea is to transfer the dc of the output back to the input while (ideally) filtering off all ac outputs from reaching the inputs, via the large LC filter, and inject a small signal into the input.

For a fully differential amplifier, you need to break both the feedback loops via 2 LC filters. The small signal is injected via the large cap of the filters, ie instead of connecting them to ground, just inject a small ac source with any dc voltage (preferably 0, but it will be ac filtered by the large cap anyway). This gives you 2 possibilities, either inject 1V ac into one of the inputs, or inject 0.5V but with 180 phase difference into both inputs.

Do note that the fully differential amplifier has the alternate CMFB loop that has to be stable as well. Just break the CMFB loop at an appropriate point just as you would for a single-ended amplifier.