The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Does the threshold voltage of MOS transistors vary with fingers?
https://designers-guide.org/forum/YaBB.pl?num=1459445167

Message started by iVenky on Mar 31^st, 2016, 10:26am

Title: Does the threshold voltage of MOS transistors vary with fingers?
Post by iVenky on Mar 31^st, 2016, 10:26am

Hi,

I would like to know if the threshold voltage of MOS transistors vary with the number of fingers in these advanced technologies. I mean compared to multipliers, does the threshold vary?

Thanks!

Title: Re: Does the threshold voltage of MOS transistors vary with fingers?
Post by loose-electron on Mar 31^st, 2016, 11:09am

As a guess, probably not. Important that you get your foundry specific data to get a meaningful answer.

Title: Re: Does the threshold voltage of MOS transistors vary with fingers?
Post by davidshw on Mar 31^st, 2016, 11:40pm

yes, dut to LOD and WPE stree effect, the vth is different of different fingers, the matching MOS devices (such as current mirror) should have the same fingers in pre-simulation.

Title: Re: Does the threshold voltage of MOS transistors vary with fingers?
Post by raja.cedt on Apr 1^st, 2016, 2:18am

Multipliers are separated by ST kind of isolation, so fingers adjacent to STI will have higher stress during manufacturing. Where as fingers are adjacent to another finger hence no stress. So Vth strongly depends on where they are.

Please refer LOD and STI effects in Google.

Best regards,
Raj.

Title: Re: Does the threshold voltage of MOS transistors vary with fingers?
Post by loose-electron on Apr 3^rd, 2016, 6:58pm

Go get empirical data to validate.

Too many times I have seen things like thins go in a different direction.

Real world data from real world silicon.

Title: Re: Does the threshold voltage of MOS transistors vary with fingers?
Post by RobG on Apr 11^th, 2016, 12:58am

raja.cedt wrote on Apr 1^st, 2016, 2:18am:

This is true, but I haven't had to lay them out that way to pass LVS. For example, if the device was M=10, NF=2, you could lay it out as M=1, NF=20 and pass LVS. Extracted simulations are real important to model this sort of stuff.

Dummy devices should be used so the surroundings are the same no matter how many fingers you use. If you use them, it won't matter if you use M or NF.