The Designer's Guide Community
Forum
Welcome, Guest. Please Login or Register. Please follow the Forum guidelines.
Jul 21st, 2024, 7:25am
Pages: 1
Send Topic Print
How to calculate k'=u*cox value in mos? (Read 9190 times)
hande.vinayak
Junior Member
**
Offline



Posts: 22
BOM
How to calculate k'=u*cox value in mos?
Apr 19th, 2007, 7:50am
 
Hi to all,
  Can anybody tell me, how to calculate k'=u*cox value in mos?
  I am using TSMC 90 nm PDK.I am working on cadence spectre for simulation.
  Plz,do the needful.
  Thnx,
Back to top
 
« Last Edit: Apr 19th, 2007, 9:38pm by hande.vinayak »  
View Profile   IP Logged
ywguo
Community Fellow
*****
Offline



Posts: 943
Shanghai, PRC
Re: How to calculate k'=u*cox value in mos?
Reply #1 - Apr 20th, 2007, 7:14am
 
Hi, hande,

I have tried to simulate the I/V curve and extract the K' value. It is impossible to calculate K' value from the model parameter because of the complexity of notorious BSIM3 model.


Thanks
Yawei
Back to top
 
 
View Profile   IP Logged
Mobeedikk
New Member
*
Offline



Posts: 4

Re: How to calculate k'=u*cox value in mos?
Reply #2 - Apr 20th, 2007, 9:00am
 
What I usually do is get a representative W/L of the type of transistor I'd use, then
extract Vth by:
1. plotting Ids vs Vgs with Vds at some small voltage (~50mV is typical).
2. take the derivative of this to get gm
3. find the Vgs that corresponds to the maximum gm
4. linearize the Ids vs Vgs curve around this Vgs point, and extrapolate back to the x axis. This will give you Vth
I then plot out an Ids vs Vds curve, and do a least squares fit of the equations to the IV curve to find k'. If you're using square law, fit the square law equation to the IV curve. If you're using Rabaey's velocity saturation equation, fit that equation to the IV curve.
Back to top
 
 
View Profile   IP Logged
Geoffrey_Coram
Senior Fellow
******
Offline



Posts: 1999
Massachusetts, USA
Re: How to calculate k'=u*cox value in mos?
Reply #3 - Apr 23rd, 2007, 4:17am
 
Hmm .. I thought there was a "ueff" output parameter, but I don't see it.  There is a betaeff in Spectre, which ought to be ueff * Cox * Weff / Leff.  Can you use that?
Back to top
 
 

If at first you do succeed, STOP, raise your standards, and stop wasting your time.
View Profile WWW   IP Logged
Pages: 1
Send Topic Print
Copyright 2002-2024 Designer’s Guide Consulting, Inc. Designer’s Guide® is a registered trademark of Designer’s Guide Consulting, Inc. All rights reserved. Send comments or questions to editor@designers-guide.org. Consider submitting a paper or model.