Designer’s Guide Community

ortali

New Member

Offline

Posts: 1

Vdsat vs VOD
Aug 12^th, 2013, 12:50am

Hi Guys, I simulate a transistor in hspice, which I want to put in saturation. I recieve vdsat = 110mV, vod = 146mV, vds=142mV. The simulator states that the transistor is in linear region.
I thought that saturation happens when vds> vdsat, however, it seems that vds also must be bigger than vod. Is this correct? It is 0.065u technology, and the L=0.5u.

Thanks a lot,
Ortal.

Back to top

IP Logged

raja.cedt Senior Fellow Offline Posts: 1516 Germany	Re: Vdsat vs VOD Reply #1 - Aug 12^th, 2013, 1:02am hello, Linear region........vgs-vth(vod)>vds....hence Linear region.
Back to top	IP Logged

Kevin Aylward Community Member Offline Posts: 52	Re: Vdsat vs VOD Reply #2 - Aug 12^th, 2013, 9:43am Comment: Vgst (vgs-vt or vod) is approximately equal to vdsat, so they get interchanged in paper and pencil analysis. Analog is all about approximations. 110mv IS ~ 140mv. Regions are not sharp. If you want full, high output impedance (low gds), you should run a few hundred mv over the alleged vgst, possibly 1/2V!
Back to top	Kevin Aylward IP Logged

harpoon

Senior Member

Offline

Posts: 112
Hampshire

Re: Vdsat vs VOD
Reply #3 - Aug 13^th, 2013, 2:32am

I am not familiar with 65nm models ... but in BSIM4, if vgs-vt(vod)>0 and vds > vdsat, you are in saturation.

however, there are 3 different saturation regions depending on vds ... namely channel length modulation (CLM), DIBL and Substrate Current Induced Body Effect (SCBE), as vds is increased beyond vdsat.

... at least that is what spectre is reporting.

I would check the hspice and process model manuals if you are still curious.

This is all academic of course ...

... Kevin's comments makes more practical sense !

Back to top

IP Logged