From: Ray Anderson ([email protected])
Date: Fri Feb 04 2000 - 13:28:44 PST
I was too quick with the send button on my
previous message. Found the problem. I didn't have
enough data points in my simulation and spice jumped
over the peak. By increasing the number of frequency
points I am now getting answers approximating yours.
It may be good to note that the values shown
in your appendix 1 example (inductance and ESR) in a couple
of cases are very, very far outside the range that would
be found in actual components. Simulations done with these
extreme values (like .00001 ohms ESR and 10nH mounted inductance)
can lead to some frightening results at first glance. Extreme
values such as 3-4K ohms are seldom seen with real world components
mounted in the appropriate manner.
A very, very good decap can have an ESR in the
10's of milliohm range, and proper mounting procedures
can lead to inductances in the sub nH range (~.4nH) These
ranges of real world validity seem to be exceeded in the
examples shown in your appendix.
I'm much more comfortable with your numbers
now that I've been able to replicate them by other means.
While it appears we may have a few differences of opinion on
some key areas, your paper is well written and points out
some important things the designer needs to be aware of.
> Date: Fri, 4 Feb 2000 12:40:05 -0800 (PST)
> From: Ray Anderson <[email protected]>
> Subject: Re: [SI-LIST] : ESR and bypass caps
> To: [email protected]
> MIME-Version: 1.0
> I ran your first test scenario from Appendix 1 in your paper
> under Hspice.
> L C R
> 10.0nH .100uF .00001
> 1.0nH .010uF .00001
> .1nH .001uF .00001
> (10 each of each capacitor value.)
> From the spice simulation I get the following results:
> FREQ IMPEDANCE (ohms)
> 5.04MHz 2.62e-5
> 15.3MHz 39.8
> 50.4MHz 2.62e-5
> 165 MHz 20.6
> 504 MHz 2.62e-5
> There is quite a discrepancy in the amplitude of the anti-resonance
> Yours: 4003 ohms Mine: 39.8 ohms
> Yours 3936 ohms MineL 20.6 ohms
> I've attached the graphic results of the spice run along with
> the little spice deck I used to create it. Any idea why we are
> seeing such a delta between the spice run and your numerical
> -Ray Anderson
> Sun Microsystems
**** To unsubscribe from si-list: send e-mail to [email protected] In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:34:59 PDT