strange transient reponse of low pass filter

[yixiusky]

I attach the ac response and transient simulation result of my low pass filter. The circuit is exactly differential and symmatric. But the transient show different voltage swing. Could you help me to explain this problem? thank you very much

(1) ac response of LPF


(2) zoom in to cutoff frequency part


(3) 3dB bandwidth


(3) transient simulation


(4) zoom in to transient simulation

 

[FvM]

The circuit is exactly differential and symmatric.

Obviously, it isn't. Either the stimulation or another circuit detail must be expected unsymmetric. Show a circuit for clarification.

 

[LvW]

I attach the ac response and transient simulation result of my low pass filter. The circuit is exactly differential and symmatric. But the transient show different voltage swing.

Why do you simulate over app. 500 periods ?
Try another simulation over app. 10-20 periods with smaller time steps, perhaps the picture looks different.

 

[yixiusky]

i attach the schematic. this is gm-c type low pass filter

this is the gm-cell used in the filter



this is the schematic of the filter


Added after 13 minutes:

Hello LvW, I tryed "10-20 periods with smaller time steps," the response still the same

 

[zopeon]

Hi yixiusky,
Have you connected your Bulk to a suitable potential (Its not clear from Figure) Check each component value (just to be sure its not a chance error, coz sometimes even if you select apply to selected only the first selected device's parameter change).

Hey why isnt there any connection between your Diff amp load's gate and the cmfb load gate? (Atleast not visible in pic).

Was your diffamp tran output proper? What is your supply and expected dc output value?
I think you are using input signal freq either outoff band or it where the gain is dependent on signal amplitude.

zopeon

 

[yixiusky]

Thank you very much zopeon.

I check everything you mentioned, they are no problem, and also the CMFB connection has no problem ^^

supply is 1.2V. DC output voltage is 720mV ^^

 

[zopeon]

Hi yixiusky,

Did you check the two outputs for the same cell in open loop? compare the same when in feedback. I think the feedback is causing different impedance at both inputs( as seen by input). Just check operating the signal at lower freq (abt 10MHz).

Hope this helps.

zopeon

 

[dick_freebird]

I don't see the input sources anywhere in the pics.
Maybe these are off-center and pushing asymmetry
through the circuit. I like to use a pair of equal
VDM/2 sources, one right-side-up and one upside-
down, both on a common VCM pedestal source for
simulating differential circuits (just to keep me from
"oopses").