Simulation with and without body effect.

[Pipeline]

when we simulate cascode amplifiers using HSpice,
With body effect: connecting body to ground
Without body effect: connecting body to source

Is it right? Thanks a lot!

 

[bastos4321]

Yes, but the normal situation in cascoding is to have the body effect. Using the bulk to source connection (no body effect) will increase the parasitic capacitance of the cascode node due to the required well to isolate bulks.

Bastos

 

[ambreesh]

If u have twin tub process u could isolate the cascode transistors. Anyway for PMOS u can do that. The body effect would increase the Vt of the cascoded device. I would only use isolation if it critical and i have low supply voltage. Also if you are designing high swing cascodes then the back bias observed is around 200mV or less and not a major effect is observed in Vt variation

 

[Pipeline]

with body effect, the small-signal gain will increase. Can this be considered as an advantage?

BTW, how can I find the r0 from hspice simulation results. Directly read from the results? If use 1/(gamma*Id), how to find gamma?

Thanks!

 

[ambreesh]

Deae pipeline,
u are talking of lambda, the channel length modulation parameter if i am not mistaken.
Hspice gives "gds" for MOS and inverse of that is R.
But lambda should be provided by the foundry, if they donot provide it, then we need to chracterise our mos.
The process is given in appendix in Allen holberg.

 

[Pipeline]

Dear ambreesh,

Thank you. I made a mistake by typing gamma instead of lambda.