Re: Shielding

Andy Ingraham (ingraham@wrksys.ENET.dec.com)
Wed, 2 Jul 97 08:31:45 EDT

The one application where I have been told guard traces actually work,
is at DC and very low frequencies, around sensitive high impedance
circuits, especially for surface etch where there may be contaminants.

As high frequency shields, their effectiveness is debatable.

What is the voltage on these guard traces? It isn't going to be 0V.
You'll have a current through a moderate trace impedance; not like the
current through a ground plane. So maybe you can reduce this voltage
by tying the guard traces to the ground plane through vias spaced
every centimeter or so (how close is close enough?). Is that really
beneficial?

There was some discussion about guard traces on this maillist within
the last year. I recall there was some empirical evidence that
removing the guard traces (and leaving open routing channels) was more
beneficial. Designers who insisted on guard traces, ended up ripping
them out on the final board rev.

Regards,
Andy