plot gm/gds vs gm/id

Status
Not open for further replies.
F

fat_123

Member level 1
Joined
Apr 11, 2020
Messages
36
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
251
hello,


i am doing gm/id technique and for that i have to plot intrinsic gain i.e gm/gds

since in MOSFET we can directly plot it.

But i am doing it for CNFET so could anyone help me out and tell me how to plot gm/gds in other devices?
 

okay so CNFET means carbon nanotube field effect transistor.

i am doing it on HSPICE
 

Show me followings.
Click to expand...

i was asked to plot gm(by sweeping vgs and vds=vdd/2) and in another file plot gds(sweeping vds and vg=vdd/2)
and them combine them.but it gave a straight line.

please suggest me something i am stuck here.
 

fat_123 said:
CNFET means carbon nanotube field effect transistor.
Click to expand...
How do you implement CNFET in HSPICE ?
What device model do you apply as CNFET in HSPICE ?

fat_123 said:
i was asked to plot gm(by sweeping vgs and vds=vdd/2)
and in another file plot gds(sweeping vds and vg=vdd/2)
Click to expand...
How do you evaluate gm and gds ?
What analysis do you use ?

I don’t think .OP is valid for CNFET.

What post processing tool do you use for HSPICE ?
AvanWave ?
CosmosScope ?
Custom Waveviewer ?
 
Last edited:

CNFET model i took from stanford

i plotted gm by sweeping vgs and set vd=vdd/2 and got the gm(right) curve

i did the dc analysis

but i am stuck for plotting gds

yes .op is not valid
 

i am using custom waveviewer
 

fat_123 said:
CNFET model i took from stanford
Click to expand...
Can you understand device model in HSPICE surely ?

What model do you use ?
Macro Model based on some FET model which HSPICE has natively ?
Compact Model based on Verilog-A description ?

fat_123 said:
i plotted gm by sweeping vgs and set vd=vdd/2 and got the gm(right) curve
i did the dc analysis
but i am stuck for plotting gds
Click to expand...
Can you understand gm and gds surely ?

You have to evaluate derivative of Id regarding Vgs and Vds, if you rely on DC analysis.

Do you evaluate derivative of Id surely ?
Derivation is a function of Custom waveviewer not HSPICE.
 
Last edited:

Stanford CNFET Model - HSPICE taken from Stanford site

32nm technology model.



yes gm=dId/dVgs gds=dId/dVds
i did the derivative of id w.r.t vgs and attached the gm plot

but how to do the same for gds and plot intrinsic gain
 

fat_123 said:
Stanford CNFET Model - HSPICE taken from Stanford site
32nm technology model.
Click to expand...
You can not understand device model for simulator at all.
32nm technology model has no relation at all.

Learn device model and simulator surely.

If you use Stanford University VS-CNFET(Virtual Source CNFET) Model,
it is a Compact Model based on Verilog-A description.

You can extract gm and gds as internal calculation result, if you can understand model equations.

fat_123 said:
but how to do the same for gds and plot intrinsic gain
Click to expand...
Surely learn and understand Custom Waveviewer.
Or use functions of ACE(Analysis Command Environment) manually in Custom Waveviewer.
Actually your operation using GUI is done by function of ACE.

If you choose other post processing tool in stead of Custom Waveviewer, you can use MATLAB or Octave with HSPICE Tool Box.
https://www.cppsim.com/download_hspice_tools.html
 
Last edited:

we cant just plot gds like gm.
i want to design opamp. for that i am doing gm/id technique and for single n-fet and p-fet plotting gm,gds,id/w
 

Show me netlist for both gm and gds.
And show me sx log file of Custom Waveview.
 
Last edited by a moderator:

Last edited by a moderator:

I checked the source code for verilog-a implementation for carbon nanotube fets and I could not find the definition of gm.
But I know other veriloga models, for example for BSIM-IMG, the operating parameters are defined.
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…