unstable with bandpass filter using active inductor

When checking AC response of bandpass filter look perfectly, put VAC=1, VDC=V0 ( suitable Bias for operating).
But when I apply transient simulation the output unstable.

what will happen if
1. changing size of Min?
2. Change Ib at buffer how big is it good.

Pls see attached file.

bicave said:

When checking AC response of bandpass filter look perfectly, put VAC=1, VDC=V0 ( suitable Bias for operating).
But when I apply transient simulation the output unstable.
what will happen if
1. changing size of Min?
2. Change Ib at buffer how big is it good.

Pls see attached file.

Click to expand...

In this case, I assume the circuit really is unstable.
Once again - as mentioned several times ago: The ac response (magnitude) alone cannot proove stability.
Did you check the ac phase response as well? In case of instability it shows the false behaviour in the vicinity of the pole frequency (rising phase instead of decrease).

I see this structure in many literature why do you assume circuit really unstalbe?
What will need to do if phase response incorrect?

I see this structure in many literature

Click to expand...

What structure do you refer to?

why do you assume circuit really unstalbe?

Click to expand...

In your first post you said that in transient analysis you noticed instability at the output signal,is this true or not?
What kind of instability,common mode or differential?
It would be a good assist to us if you show us your testbench with details for transient analysis and a plot of the output signal.

Additionally what kind of filter is the one you design?Active-RC,OTA-C...?

What will need to do if phase response incorrect?

Click to expand...

If you indeed designed an unstable filter,you can take some information from the phase response as LvW said and a pole-zero analysis of your circuit
will show a definetely good image of what is really going on.
For example,you may observe from PZ analysis that you have poles in the RHP so you must find a method to move them to LHP or
perform some kind of compensation.
In another case you may have instablity if you inject a common mode signal in your filter.In this case you must investigate what is going at the inside structure of the filter (at the opamp let's say for an Active-RC implementation or at the buffers etc.)

Provide more info to assist you further.

What is VI and VQ?

bicave said:

When checking AC response of bandpass filter look perfectly, put VAC=1, VDC=V0 ( suitable Bias for operating).
But when I apply transient simulation the output unstable.

what will happen if
1. changing size of Min?
2. Change Ib at buffer how big is it good.

Pls see attached file.

Click to expand...

how did you decide that V0 was a suitable operating point?
did you try performing a proper stability analysis on the feedback loop instead of an AC analysis?
the amount of current in the output follower is dictated by the maximum slew on the output waveform, for a sinusoidal waveform this depends on frequency and amplitude
how did you tune the variable caps?

In fact, I didn't see the completed feedback loop.
You should show us full picture for stability analysis.

bicave said:

I see this structure in many literature why do you assume circuit really unstalbe?
What will need to do if phase response incorrect?

Click to expand...

That's funny.
I didn't "assume" instability. In contrary, it was your own statement (if you remember your 1st posting):
"But when I apply transient simulation the output unstable.
"
I only wanted to point to the fact that ac magnitude analysis alone is not sufficient.

My aim to design active filter using active inductor but I don't see filter characteristic at all at time domain, by putting a sin wave with sin wave , properly DC operating, and amplitude of 10mV for example.
The attached file is schematic & ZIN ( check ZIN to ensure active inductor behave like RLC equivalent circuit)





How can I configure port to plot S21 and S11 of filter? I really don't know why filter don't have correct behavior in time domain

A picture of your filter's testbench is better than words.Show it in details (biquads,input biasing,setup for transient etc...)
Also give us a plot of your input (magnitude of the signal and frequency with markers) and output signal (magnitude of the signal and frequency with markers as well) at time domain in order to see where is the problem.
The above images say nothing about the instability you mentioned previously in your posts...

My aim to design active filter using active inductor but I don't see filter characteristic at all at time domain, by putting a sin wave with sin wave , properly DC operating, and amplitude of 10mV for example.

What do you expect in the time domain to confirm the filter characteristic?
Didn't you check frequency transfer response ?

LvW said:

My aim to design active filter using active inductor but I don't see filter characteristic at all at time domain, by putting a sin wave with sin wave , properly DC operating, and amplitude of 10mV for example.

What do you expect in the time domain to confirm the filter characteristic?
Didn't you check frequency transfer response ?

Click to expand...

1. I expect when out of filter band, signal @ output of filter attenuate. ie 1GHz is frequency out of bandwidth, out put signal expect 1mV for example due to magnitude respose of filter.
2. Yes, I use a voltage short with Vdc=V1, VAC=1 V and sweep Vdc to check the frequency respose. V1 is chosen when the shape of filter meet expect.
Filter frequency response should be similar with Impedance frequency response of Active inductor.

bicave said:

2. Yes, I use a voltage short with Vdc=V1, VAC=1 V and sweep Vdc to check the frequency respose. V1 is chosen when the shape of filter meet expect.
Filter frequency response should be similar with Impedance frequency response of Active inductor.

Click to expand...

Voltage short? Sweep Vdc for frequency response? ...should be similar with response of inductor ?

I really feel lost. When I ask for a frequency response I ask for RESULTS not "how it should be"!

LvW said:

Voltage short? Sweep Vdc for frequency response? ...should be similar with response of inductor ?

I really feel lost. When I ask for a frequency response I ask for RESULTS not "how it should be"!

Click to expand...

Sorry "voltage source"