RE: [SI-LIST] : on-chip decoupling capacitance (and SI)

About this list Date view Thread view Subject view Author view

From: Chris Cheng ([email protected])
Date: Wed Apr 19 2000 - 12:29:04 PDT


only partially true, in particular with deep submicron device.
trench isolation on wells tends to have very high resistance
and very little control on variation. u will be significantly
off if u base on esr purely on area resistance.
chris

-----Original Message-----
From: D. C. Sessions [mailto:[email protected]]
Sent: Wednesday, April 19, 2000 11:58 AM
To: [email protected]
Subject: Re: [SI-LIST] : on-chip decoupling capacitance (and SI)

"Zabinski, Patrick J." wrote:
>
> > > > Case in point: we just ran the onchip capacitance extraction
> > > > for a customer
> > > > device. 200nm technology, before adding intentional bypass
> > > > devices, the
> > > > assorted sources of supply capacitance came to ~230 nF in the
> > > > 100-1000 MHz
> > > > range.
> > >
> > > D.C.
> > >
> > > Can you shed some light as to what you mean by "assorted
> > > sources of supply capacitance"? It's hard for me to believe
> > > 230 nF of trace/grid-capacitance, so I'm struggling to figure
> > > out where it's coming from.
> >
> > Most of your logic is quiescent. RAMs especially. And when
> > you get down to it, that means that you have a great deal
> > of gate-source capacitance in series with fully-on source-drain
> > paths. The net effect is a truly enormous amount of charge
> > reservoir, damped by parallel high-ESR drain-well and
> > well-substrate capacitance.
>
> D.C.,
>
> Thanks; this helps.
>
> I can understand and reasonably-estimate the effects of the "gate-source
> capacitance in series with fully-on source-drain paths" without
> too much difficulty with Spice. However, I have yet to see an effective
> model within Spice/Hspice that acocunts for the "high-ESR
> drain-well and well-substrate capacitance"; how do you
> determine the effective capacitance for these well-capacitances
> and the associated ESR?

Wet-thumb triangulation from a couple of points. First off, the
well-substrate junctions are very well understood -- we have excellent
control over their doping profiles and know to a fine degree what their
areal capacitance is. Secondly, we can measure it on testers at low
frequencies. Finally, we don't give a flip what the exact values are
so long as we can run plausible back-of-the-envelope calculations to
convince ourselves that there's enough loss in the system to keep the
Q managable.

-- 
D. C. Sessions
[email protected]

**** To unsubscribe from si-list or si-list-digest: send e-mail to [email protected]. In the BODY of message put: UNSUBSCRIBE si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu ****

**** To unsubscribe from si-list or si-list-digest: send e-mail to [email protected]. In the BODY of message put: UNSUBSCRIBE si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu ****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:36:14 PDT