Designer’s Guide Community

zhonghan Junior Member Offline Posts: 16 china	what is the spectre format of mos transistors ? &n Mar 20^th, 2006, 1:42am i just want to write a simplest spectre model by myself. I know the format in Hspice , like .model mn nmos vth0=0.7 gamma=0.4 lambda=0.05. what is the coresponding format in spectre( incuding version )? Then how to cite this model in spectre , I know " .lib 'path' corner " in HSPICE is just ok .
Back to top	IP Logged

Geoffrey_Coram

Senior Fellow

Offline

Posts: 1999
Massachusetts, USA

Re: what is the spectre format of mos transistors
Reply #1 - Mar 20^th, 2006, 4:15am

Spectre can parse Spice syntax, so you should be able to use the HSpice-style .model card -- particularly if you're using the .lib statement, which is also Spice.

We usually use Spectre format:

simulator lang=spectre
model mn mos1 type=n vto=0.7 gamma=0.4 lambda=0.05

What LEVEL are you using? vth0 is a BSIM3/4 parameter (level=1 uses vto), but gamma is a level=1 or level=3 parameter (bsim3v3 has gamma1, gamma2 but no gamma).

Back to top

If at first you do succeed, STOP, raise your standards, and stop wasting your time.

IP Logged

zhonghan Junior Member Offline Posts: 16 china	Re: what is the spectre format of mos transistors Reply #2 - Mar 20^th, 2006, 7:36am Thank you , I have resolved this issue. you are right , I just copy the "vth0 " from a bsim3v3 model. By the way ,do you know the simple way to calculate the Ro of transistors with a bsim3v3 model?
Back to top	IP Logged

Geoffrey_Coram Senior Fellow Offline Posts: 1999 Massachusetts, USA	Re: what is the spectre format of mos transistors Reply #3 - Mar 22^nd, 2006, 9:18am There's not an easy way directly from the parameters, but I think you get it from the slope of the Id/Vd curve (or 1/Gds for a Vd sweep).
Back to top	If at first you do succeed, STOP, raise your standards, and stop wasting your time. IP Logged

zhonghan Junior Member Offline Posts: 16 china	Re: what is the spectre format of mos transistors Reply #4 - Mar 24^th, 2006, 11:07pm I know the 1/ro is equal to the slope of the Id/Vd curse. But the ro is dependent on the operating point, in the other word dependent on the Vgs. So does it mean I should get different ro under different Vgs ?
Back to top	IP Logged

Geoffrey_Coram Senior Fellow Offline Posts: 1999 Massachusetts, USA	Re: what is the spectre format of mos transistors Reply #5 - Mar 27^th, 2006, 3:26am Yes, you should. I suppose, if you are in saturation, then the Id/Vd curves are nearly parallel for different Vgs, so ro will not be strongly dependent on Vgs.
Back to top	If at first you do succeed, STOP, raise your standards, and stop wasting your time. IP Logged