RE: [SI-LIST] : Parallel Plate Capacitance for Bypass

About this list Date view Thread view Subject view Author view

From: Greim, Michael ([email protected])
Date: Mon Mar 27 2000 - 11:23:28 PST


Coupla things.......

You appear to be off by a factor of 2 somewhere. Plane
capacitance is defined as

        Cplane = (0.225*Er*A)/d

for a 2 mil core, with A being 1 square inch, this becomes

        Cplane = 0.225*4.1/.002 = roughly 460 ohms

With standard FR4 you can run into a couple of problems
that you might want to check on. Thin core materials can
prove to be quite costly and i have heard that during laminating
that all the resin can be squeezed out creating air pockets.
Therefore, before going with a standard material, you might
want to consider the benefit to cost tradeoffs of materials that
are targetted for this purpose such as Emcap or BC2000.
I found good information on these materials on the Hadco and
Merix web sites.

I would also recommend running a frequency simulation on
your PDS to make sure that you don't have frequency regions
that have very high impedances. Suffice it to say that bad
things will occur in your system if this occurs. There are a
number of good resources (most notably and referenced,
those by Larry Smith et al.) available on the web to help
you with your design. You might also want to check out
Istvan Novaks website as he has a great paper and utility
for general download. He can be found at:

        home.att.net/~istvan.novak/

Best of Luck,

Michael Greim

> -----Original Message-----
> From: Ron Miller [SMTP:[email protected]]
> Sent: Monday, March 27, 2000 1:39 PM
> To: [email protected]
> Subject: Re: [SI-LIST] : Parallel Plate Capacitance for Bypass
>
> Thanks Steve, I knew it was somewhere around there.
>
> So, by extension, 2.2 mil dielectric, available at very little if any
> extra cost should be about 1000pf/in sq for FR4
> and eliminatee the need for discrete capacitors below 1000 pf.
>
> ron
>
> "S. Weir" wrote:
>
> Ron,
>
> 5 mils should yield about 200pF / in sq, before taking away for
> vias.
>
> Regards,
>
>
> Steve.
> At 02:30 PM 3/24/2000 -0800, you wrote:
>
> Chris
>
> Go to the AVX webpage and the American Technical Ceramics
> web
> pages. They have S-parameters which show that the smaller
> profile
> ceramics are good to above 1 Ghz.
>
> Small profile usually means smaller capacitance so you may
> need
> a small one of .01 uf and a larger one of about .47 uf.
>
> For frequencys 3 Ghz and up use a capacitive layer in your
> board. Recently I heard that with 5 mil FR4 dielectric you
> get about
> 100pf per square inch. So, using a 2 mil layer would give
> about 7.5 X
> this value(per square inch).
>
> Ron Miller
>
>
>
>
> According to specifications, X7R type chip
> capacitors used for power subsystem bypass seem to become ineffective
> above 150 to 200 Mhz. For frequencies higher up, you are reliant upon the
> internal parallel plate capacitance. I have two questions. 1. Is the
> calculation for the required capacitance that is needed for the parallel
> plate capacitance the same as that used for a discrete chip capacitor
> bypass network, or is there a conservation of charge situation where the
> real answer is Cpp*dVnoise = Cload*dVchange? (where Cpp = parallel plate
> capacitance, dVnoise = tolerable change in power supply voltage, Cload =
> sum of load capacitance being switched simultaneously, dVchange = voltage
> change through driver output switch). 2. If you have a power plane
> sandwiched between 2 ground planes, does the parallel plate capacitance
> double for a given area?Thank you,Chris HansenSr. Design EngineerDPT &
> Adaptec Companies
>
> --
> Ronald B. Miller _\\|//_ Signal Integrity Engineer
> (408)487-8017 (' 0-0 ')
> fax(408)487-8017
> ==========0000-(_)0000===========
> Brocade Communications Systems, 1901 Guadalupe Parkway, San
> Jose,
> CA 95131
> [email protected], [email protected]
>
>
> --
> Ronald B. Miller _\\|//_ Signal Integrity Engineer
> (408)487-8017 (' 0-0 ') fax(408)487-8017
> ==========0000-(_)0000===========
> Brocade Communications Systems, 1901 Guadalupe Parkway, San Jose, CA
> 95131
> [email protected], [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
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:35:56 PDT