nmos pmos ro is too low for the Gain result

Thread Starter

yef smith

Joined Aug 2, 2020
717
Hello,I have build a common source with 28.5dB gain which is 25.11 linear gain in LTSPICE.

From gm/id value:
Id = 50 μA gm/id = 12.6 => gm = 12.6×50×10-6 = 630×10-6
Id vg Vds in the Vgs conditions is shown below.
g_ds = d_Id/d_Vds plot is shown in the end.

For Vds = 0.9 I get rds_nmos = 1.9×10-5. These results are for NMOS. A = gm×R=(1.9×10-5)×(630×10-6) which is far away from 25.11. In the end i have added pmos d_I/d_Vds i get there g_ds=0.0001
from equation shown in the end i get 1/gds_n=52085 1/gds_p=4887 still A=gm*Rt=4.41 far away from 25
Where did I go wrong recreating the simulation result?
Thanks.

1680278798627.png

1680278817942.png

1680278837256.png

1680278857004.png

1680278863648.png
1680278881771.png

1680278902233.png

1680278923358.png
 
Last edited:

Thread Starter

yef smith

Joined Aug 2, 2020
717
Hello crutschow,My biasing is great as you can see bellow.i get 50uA and my deisgn is working fine.
The problem is that my ro calculation is about 4K ohms
i have made gds=d_I/d_vds
could you say where did i go wrong in this aspect?
Thanks.
1680283237492.png

1680283362836.png
 

crutschow

Joined Mar 14, 2008
34,285
Since I don't understand why you are biasing a P-MOFSET with a positive drain voltage, I can offer no further relevant comment.
 

Thread Starter

yef smith

Joined Aug 2, 2020
717
UPDATE:

Hello crustchow, from the p-channel mosfet laws.
Vs=1.8 Vg=0.891 Vd=0.9 Vt=-0.4
Vsg=1.8-0.891>|-0.41| so its ON

Vsd=1.8-0.9=0.9> Vsg-|Vt|=0.509
So its conducting and in saturation state.
I dont see any problem .
Its not what i have asked in the original post but 'just to explain you my logic which is also perfectly working in the simulations of the PMOS shown above.
 

Thread Starter

yef smith

Joined Aug 2, 2020
717
UPDATE:

Hello crustchow, from the p-channel mosfet laws.
Vs=1.8 Vg=0.891 Vd=0.9 Vt=-0.4
Vsg=1.8-0.891>|-0.41| so its ON

Vsd=1.8-0.9=0.9> Vsg-|Vt|=0.509
So its conducting and in saturation state.
I dont see any problem .
Its not what i have asked in the original post but 'just to explain you my logic which is also perfectly working in the simulations of the PMOS shown above.


Could you please help me with the ro?
 

crutschow

Joined Mar 14, 2008
34,285
My confusion is that you are showing the bottom transistor with a P-MOSFET symbol, whereas the publication shows an N-MOSFET, so which are you using?
(Note that there are two different types of MOSFET symbols being used, which may be the source of the misunderstanding. Look at which terminals the arrows are connected).
 

Thread Starter

yef smith

Joined Aug 2, 2020
717
Hello AudioGuru, its a validation issue.
I want to match my gain result with theory.
A=gm*(ro_n||ro_p)

I am using a level 8 spice model shown bellow.
gm = 12.6×50×10-6 = 630×10-6 because
g_ds_pmos=0.0001845 => rds_pmos=5420
I get r_ds very small from regular derivative.
How can i get my ro from the spice model?

Spice model:
https://sanjayvidhyadharan.in/Downloads/tsmc_180_nm/tsmc018.lib

T58F SPICE BSIM3 VERSION 3.1 PARAMETERS
*
* SPICE 3f5 Level 8, Star-HSPICE Level 49, UTMOST Level 8


Spice model manual:
https://class.ece.uw.edu/cadta/hspice/chapter_16.pdf
1680336261673.png
1680336818193.png
1680336705898.png

1680338467455.png
 
Top