The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> How deep in Subthreshold region can we drive a MOSFET?
https://designers-guide.org/forum/YaBB.pl?num=1464866357

Message started by jdp on Jun 2^nd, 2016, 4:19am

Title: How deep in Subthreshold region can we drive a MOSFET?
Post by jdp on Jun 2^nd, 2016, 4:19am

In an opamp, we can set the input-pair MOSFETs to be in Sub-threshold region of operation.
To what extent can we put them in sub-threshold?...like, can we use (Vgs - Vth) negative?
As I searched in forum posts, people say they use (Vgs - Vth) ~ 100-200mV which makes the MOSFET near subthreshold..but is it feasible to use a Vgs somewhat < Vth ?

(I would be using the opamp in a lower freq. application (say max. 5 MHz))

Title: Re: How deep in Subthreshold region can we drive a MOSFET?
Post by loose-electron on Jun 2^nd, 2016, 1:37pm

what does your simulation model tell you? Is the model valid at that range of bias? You need to examine the vailidity of the model to get a real answer.

http://chipdesignmag.com/display.php?articleId=438

give that a read

Title: Re: How deep in Subthreshold region can we drive a MOSFET?
Post by jdp on Jun 8^th, 2016, 9:38am

Thanks loose-electron for your reply :) I get what you suggest...

By the way, if anybody has some relevant information/experience (how much can Vgs be made < Vth in Subthreshold) regarding UMC 180nm process might share...

Regards.

Title: Re: How deep in Subthreshold region can we drive a MOSFET?
Post by rfmagic on Jun 15^th, 2016, 2:38pm

The first place to look is the model-reference-manual provided by the foundry. The manual should include the measurements that were taken from various test structures and you can verify that the region you intend to work in is valid or covered by the model.
Note that the measured test structures are defined over limited set of dimensions and bias points so pay attention to that.

Title: Re: How deep in Subthreshold region can we drive a MOSFET?
Post by RobG on Jun 22^nd, 2016, 11:19pm

jdp wrote on Jun 2^nd, 2016, 4:19am:

Sorry for the late response... you can make it as wide as you want and make Vgs-Vt negative, but at some point the leakage will really affect performance especially at hot temperature. Also, the gate to source capacitance gets larger.

I settled on Vdsat=150mV a few years ago as a reasonable starting point for a differential pair. I'm not sure what Vgs-Vt is for that operating point as it is slightly different than Vdsat, but I remember driving it negative before. Reducing Vdsat further by making the device wider has diminishing returns in gm, but the Cgs cap increases linearly.