RE: [SI-LIST] : Decoupling:routing

Grasso, Charles (GrassC@LOUISVILLE.STORTEK.COM)
Mon, 11 May 1998 10:44:18 -0600

At the risk of adding more information to this...

I have research information that shows that the addition of a ferrite
bead
in series with the VCC pin of an IC does dramatically reduce the DM
noise produced by the chip - hence reducing the noise seen
differentially
across the planes. But the measured data showed that there was NO
reduction in the CM at all.

So, the addition of a cap using traces to the planes adds a small amount
of inductance
in series with the cap and hence a low pass filter. This may explain the
improvement
seen in the analog example. This technique, I suggest, should not be
used
for decoupling digital components as the signal integrity/emissions from
the
chip will be compromised. For digital tie the cap & the chip directly to
the planes.

As Winston Churchill once said about decouplers:

Never , in the field of engineering, has so much been written by so many
-
about so few.

Thank you
Charles Grasso
EMC Engineer
StorageTek
2270 Sth 88th Street
Louisville CO 80027 MS 4262
grassc@louisville.stortek.com
Tel:(303)673-2908
Fax(303)661-7115
Symposium Website URL: http://www.ball.com/aerospace/ieee_emc.html

>----------
>From: Jay Diepenbrock[SMTP:jaydiep@us.ibm.com]
>Sent: Wednesday, May 06, 1998 7:18 AM
>To: si-list@silab.Eng.Sun.COM
>Subject: Re: [SI-LIST] : Decoupling:routing
>
>Regarding Tom Zimmerman's posting about the same subject, our experience is
>similar for mixed-signal designs. We have found that both noise within the
>system and EMI are improved by providing the best possible common Ground for
>all subsystems. Problems with digital noise getting into the analog
>circuitry
>were solved by generous use of ferrite beads and decoupling on local power
>supply connections, but we had to add local voltage regulators on the analog
>power supply lines in some cases because the noise generator (it was an
>intentional transmitter) was just so strong that filtering wasn't sufficient
>to
>keep it out. Splitting the Ground plane with common attachment back at the
>power supply was the initial approach, but the Ground noise in the analog
>circui
>try with that arrangement was too high.
>
>Jay Diepenbrock
>
>Advisory Engineer
>Interconnect Technology & Qualification
>IBM Global Procurement, B8UA/061, RTP, NC
>Phone: 919-543-8804 Fax: 919-543-3642
>Email: jaydiep@us.ibm.com
>