The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> Circuit Simulators >> Spectre dc operating points/model calculation
https://designers-guide.org/forum/YaBB.pl?num=1160016166

Message started by reilly on Oct 4^th, 2006, 7:42pm

Title: Spectre dc operating points/model calculation
Post by reilly on Oct 4^th, 2006, 7:42pm

Hi,
I'm miserable on trying to do a hand calculation during design. Most of the time, we try to start determine the proper W/L for a transistor using Id=k/2(W/L)(Vgs-vth)^2 depending the operating region desire to operate at. The problem arise here is, in the model file, the k or betaeff output from spectre is a changing value depending on the W/L of the transistor (or based on model selector). I did tried to make a thorough calculation based on bsim3v3 equation to calculate vth, vdsat etc refering to manual from cadence and bsim, yet still some mismatch arise, is it I miss out the homotopy calculation that simulator did? Can anyone point me out how spectre come out with these values for dc operating points, equation should be great if anyone can share it here. The purpose of matching the value is to aids in first hand calculation to determine the approximate W/L which I feel will increase my confidence on the dimension I had chosen and also countercheck on the result. This is the way designer should work. Any advice to offer? Thanks in advance!
signal OP("/I0/NM9" "??")
betaeff 414.964u
cbb 133.768f
cbd 7.22042a
cbg -48.4775f
cbs -85.2974f
cdb -75.4017a
cdd 3.80463f
cdg -3.93929f
cds 210.058a
cgb -56.412f
cgd -3.76909f
cgg 221.473f
cgs -161.292f
cjd 10.1642f
cjs 12.3906f
csb -77.2802f
csd -42.7591a
csg -169.057f
css 246.38f
gds 17.2005n
gm 22.5107u

Title: Re: Spectre dc operating points/model calculation
Post by Andrew Beckett on Dec 13^th, 2006, 8:35am

It's not practical for designers to do detailed hand calculations of the entire model equations these days. I don't know why anyone would want to do that. Sure, you can use the level 0 or level 1 type equations to give you the ability to judge relative relationships and approximate starting points for your circuit, but really a circuit simulator is necessary in order to check things.

You should be able to use the spectre documentation and the BSIM documentation on the BSIM site - but I seriously doubt that doing this is something that designers _need_ to do. When I was doing design, I used to use the textbook type equations to get the relationships, but leave it to the simulator to get a more realistic complete circuit picture.

Regards,

Andrew.