Designer’s Guide Community

Larry_80 Community Member Offline Posts: 61	*How does vgs-vth for optimum (gm/Id)ft vary as technology gets smaller** May 08^th, 2012, 9:14pm Hello All, Does the vgs-vth for optimum FOM = (gm/Id)*ft get larger as technology size decreases? And also, do i have to characterize this FOM for both NMOS and PMOS or i can just do for one and use for sizing my transistors? Thanks in advance
Back to top	IP Logged

vivkr

Community Fellow

Offline

Posts: 780

Re: How does vgs-vth for optimum (gm/Id)*ft vary as technology gets smaller
Reply #1 - May 10^th, 2012, 5:17am

Well! I would expect the Vgs-Vth optimum numbers to scale down as well because there is a dependence on "beta", the current gain factor which scales with technology.

However, I am not sure if you can actually operate your FET at the optimal Vgs-Vth. I noticed already in very old technologies that the current density gets so high at this point that you would have all sorts of problems getting the required amount of current into the transistor.

In older processes, people sometimes looked away if they were doing this for a handful of FETs. In the new ones, electromigration might make or break your design. At any rate, you will probably need a customized layout of your device.

Why do you ask by the way? Are you trying to decide on the technology to use?

Vivek

Back to top

IP Logged

RobG Community Fellow Offline Posts: 570 Bozeman, MT	*Re: How does vgs-vth for optimum (gm/Id)ft vary as technology gets smaller Reply #2 -** May 10^th, 2012, 10:06am Doesn't the optimum gm/Id occur for small Vgs-Vt?, i.e. lowest current density? The problem I run into is leakage current, especially at hot temperatures.
Back to top	IP Logged

ACWWong Community Fellow Offline Posts: 539 Oxford, UK	*Re: How does vgs-vth for optimum (gm/Id)ft vary as technology gets smaller Reply #3 -** May 10^th, 2012, 8:29pm Hi Rob, I think you missed the *ft part of the FOM.
Back to top	IP Logged

vivkr

Community Fellow

Offline

Posts: 780

Re: How does vgs-vth for optimum (gm/Id)*ft vary as technology gets smaller
Reply #4 - May 14^th, 2012, 5:53am

and I missed the (gm/Id) part of the question, looking only at fT...

I guess if you are trying to look at both (gm/Id) which gets maximized for low Vgs-Vth and fT, which gets maximized at higher Vgs-Vth, then you might end up with reasonable Vgs-Vth at the end. So my comment about it being difficult to drive in that much current may not hold.

As to whether the optimum scales with technology, I would expect so for the reasons outlined earlier.

Vivek

Back to top

IP Logged

RobG Community Fellow Offline Posts: 570 Bozeman, MT	*Re: How does vgs-vth for optimum (gm/Id)ft vary as technology gets smaller Reply #5 - May 17^th, 2012, 9:57am ACWWong wrote on May 10^th, 2012, 8:29pm:** Hi Rob, I think you missed the *ft part of the FOM. I did miss that. Thanks.
Back to top	IP Logged

ACWWong

Community Fellow

Offline

Posts: 539
Oxford, UK

Re: How does vgs-vth for optimum (gm/Id)*ft vary as technology gets smaller
Reply #6 - May 18^th, 2012, 9:18pm

vivkr wrote on May 14^th, 2012, 5:53am:

and I missed the (gm/Id) part of the question, looking only at fT...

I guess if you are trying to look at both (gm/Id) which gets maximized for low Vgs-Vth and fT, which gets maximized at higher Vgs-Vth, then you might end up with reasonable Vgs-Vth at the end. So my comment about it being difficult to drive in that much current may not hold.

As to whether the optimum scales with technology, I would expect so for the reasons outlined earlier.

Vivek

My experience in the technologies its usually at a quite low overdrive, but as you say its technology dependant, but also dependant to the particular device type and sizes you're looking to use.

Back to top

IP Logged