The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Modeling >> Semiconductor Devices >> MOS ft calculation
https://designers-guide.org/forum/YaBB.pl?num=1068838775

Message started by raz on Nov 14^th, 2003, 11:39am

Title: MOS ft calculation
Post by raz on Nov 14^th, 2003, 11:39am

Can someone suggest a quick and simple way to calculate ft for a MOS transistor using available op-point parameters. I am using BSIM3V3 models in Spectre version 5. It does not provide ft in the operating point information. I tried to estimate it using gm, Cgd,Cgs and Cgb but apparently these capacitances are not physical capacitance, rather they are partial derivatives so they cannot be used directly.

raz

Title: Re: MOS ft calculation
Post by Geoffrey_Coram on Feb 13^th, 2004, 10:58am

Raz-
You should be OK with gm/(2*pi*cgg)

It looks like cgg includes the overlap capacitance as well as the channel capacitance.

-Geoffrey

Title: Re: MOS ft calculation
Post by Marek Mierzwinski on Feb 15^th, 2004, 2:58pm

I don't know how easy this is in Spectre, but since fT is defined as where current gain goes to unity, plotting |Y21| vs freq will give you a very good estimate (assuming the model is accurate).

Title: Re: MOS ft calculation
Post by Geoffrey_Coram on Jun 6^th, 2005, 8:32am

Another possibility is the formula (from Gummel?):

fT = f / Imag(1/h_fe)

where h_fe is the current gain, i_out/i_in. Bias with a gate voltage with an ac magnitude, then measure the gate and drain currents. The simulation frequency "f" only needs to be reasonably close to the actual fT.

Title: Re: MOS ft calculation
Post by crhu on Oct 25^th, 2005, 5:07pm

Hi, there,
I think that we need to clarify which ft you are interested: device(ft) or block(Ft)

All the parameters(Y, H, S, Z, G) depend on the termination, as well as the device. If there is large difference between the device's input impendance and the termination impendance, the Y/H/S/Z/G are dominated by the reflection in the interface, and the Ft calculated from unity current gain, i.e.H21(db)=0 or H21(mag)=1, is far less than ft calculated from gm/(2*pi*Cg,tot).

-ch