Why there are great difference between -3dB frequency...

[Pipeline]

when I simulated a cascode amplifier, with body effect and without body effect, the -3dB frequency are much different, as 30kHz and 130kHz.
Could anybody tell me why it happens?
Since the output resistance changed 20%, and output capacitance did not have much chage, I do not understand why the -3dB frequency are much different.

 

[pico]

I think it affect yhe Q factor so much ?

 

[Humungus]

Besides the change of the output impedance, you have also gm of the input stage changed. This makes the ft and the gain to change. If these changes do not justify all the drift of your bandwidth, you coul also check if all your transistors remain saturated when you activate the body effect. Also the gate-bulk and drain-bulk capacitors are now more important, so this adds more capacitance on the output node.

 

[Vamsi Mocherla]

Body effect drastically effects the transconductance of the input differential pairs.Hence your GBW will be effected drastically. Remember gm = K W/L (VGS -VTH).

Added after 3 minutes:

Also, your output resistance is increased by the current reduction. Hence, the dominant pole will move in. I do not think that any parasitic cap will be effected drastically except the source-bulk capacitance(which is either shorted when bulk is tied to source).

Importantly, check whether your transistors are in saturation. I do not think, it will effect by that much like Humungus said......

 

[Pipeline]

For a dual cascode amplifier consistiing of 2 NMOS(M1,M2) and 2PNOS(M3,M4).
When we calculate the output capacitance, should we consider the Cdb of M2 and M3 if without body effect?
I use CAPTAB in HSPICE to list the capacitors, I see for the output node capacitance there is no big difference between without and with body effect.

Thanks.