How WPE and STI affect the mobility which is reflected in Spice or Spectre simulator?

[Alex Liao]

Hi,

The two layout dependent effects that cause mobility variation is WPE and STI. The equations are given as follows:
WPE: (Ueff,wpe)

STI: (Ueff,STI)


Beta is the parameter found after simulation and is related to Ueff(effecitive mobility).
Beta = Ueff,final * Coxe * Weff/Leff.

This Beta equation is used in simulator for sure. Is there anyone who knows the relationship between Ueff,final and Ueff,wpe/Ueff,STI?
I know the variance of mobility (Ueff,wpe/STI) is part of Ueff. But is there any equation that clearly describe the relationship between those partial mobility and the final mobility used in simulation?

 

[erikl]

... is there any equation that clearly describe the relationship between those partial mobility and the final mobility used in simulation?

I (also) didn't find such an equation in the BSIM470 Manual, but I think you can serialize the effects of WPE and STI (i.e. their ratios to µ_eff,org) :

µ[SUB]eff[/SUB](WPE) / µ[SUB]eff,org[/SUB] = (1+KU0WE∙(SCA+WEB∙SCB+WEC∙SCC))                            (15.1)

µ[SUB]eff[/SUB](STI) / µ[SUB]eff,org[/SUB] = ( ... )                                                    (14.5)

µ[SUB]eff[/SUB](WPE+STI) / µ[SUB]eff,org[/SUB] = (µ[SUB]eff[/SUB](STI) / µ[SUB]eff,org[/SUB]) ∙ (µ[SUB]eff[/SUB](WPE) / µ[SUB]eff,org[/SUB])

Of course there are a lot more effects which influence the mobility factor, like, e.g., carrier velocity saturation (mobility reduction due to the longitudinal field), vertical field mobility reduction (VFMR), and, last not least, temperature.

 

[Alex Liao]

Hi Erikl,

So 1. you think Ueff,org is U0 from model card, right?
2. do you think the Mobility Coulomb Scattering Model is such 'the others' that influence the mobility and should have the serialization relationship with previous STI and WPE factor?
3. I am almost agree with you about the serialization relationship but I am currently lack of documents that clearly prove this as well as you, but have you done any experiments to attempt to prove this?

Regards,
Alex

 

[erikl]

1. you think Ueff,org is U0 from model card, right?

Yes!

2. do you think the Mobility Coulomb Scattering ModelView attachment 101211 is such 'the others' that influence the mobility and should have the serialization relationship with previous STI and WPE factor?

I think so. But just consider such effects which actually apply in a special application.

3. ... have you done any experiments to attempt to prove this?

No. Not knowing of a better method, I'd just think this might result in a good (1^st order) approximation.

If this is really important for you, I'd suggest to try and contact one of the BSIM authors directly, e.g. Xuemei (Jane) Xi , who's e-mail address is explicitely mentioned in the BSIM4.5.0 Enhancements document: JaneXi@eecs.Berkeley.EDU .

 

[Alex Liao]

Hi Erikl,

Thanks for supporting this. I tried to contact BSIM group by leaving message on their web main page. They never reply. Also I tired to contact Xuemei at: janexi@eecs.berkeley.edu. But the mail system says

The original message was received at Mon, 20 Jan 2014 12:52:20 -0330
from mail-ob0-f169.google.com [209.85.214.169]

----- The following addresses had permanent fatal errors -----
<janexi@eecs.berkeley.edu>
(reason: 550 unknown user)

----- Transcript of session follows -----
... while talking to mx.berkeley.edu.:
>>> RCPT To:<janexi@eecs.berkeley.edu>
<<< 550 unknown user
550 5.1.1 <janexi@eecs.berkeley.edu>... User unknown



Reporting-MTA: dns; ellington.ucs.mun.ca
Received-From-MTA: DNS; mail-ob0-f169.google.com
Arrival-Date: Mon, 20 Jan 2014 12:52:20 -0330

Final-Recipient: RFC822; janexi@eecs.berkeley.edu
Action: failed
Status: 5.1.1
Remote-MTA: DNS; mx.berkeley.edu
Diagnostic-Code: SMTP; 550 unknown user
Last-Attempt-Date: Mon, 20 Jan 2014 12:52:23 -0330

It seems they decline email from outside. Do you have an alternative candidate to contact?

 

[erikl]

... I tired to contact Xuemei at: janexi@eecs.berkeley.edu.

I guess Xuemei (Jane) has left the BSIM group.

Do you have an alternative candidate to contact?

You could try this general address: bsimgroup@gmail.com

Or try one of these addresses found here.

 

[Alex Liao]

Hi Erikl,

I have sent each of the group member an email. Let's see what's happened. By the way, I saw your occupation filed: on your own. You are doing IC design on your interest? I guess you must be a professor in a university in Germany years ago.

Alex

 

[erikl]

Hi Alex, pls. see your EDAboard private mail!

Good luck, erikl

 

[Alex Liao]

Hi erikl,

One member of BSIM group have confirmed the multiplication relationship (serialization)

Ueff,final = U0 * Kwpe * K sti * K others

, Thanks.

Alex

 

[erikl]

One member of BSIM group have confirmed the multiplication relationship (serialization)

Thanks a lot for your feedback, Alex!