From: Doug McKean ([email protected])
Date: Fri Jun 23 2000 - 11:45:22 PDT
Well, I need to ask to have what I think is going on
with guard traces.
The ends of the guard trace are of course grounded.
In a monopole or dipole, either one or both ends are
high resistance and the driving point is low impedance.
This isn't the case with a guard trace. It's a low
impedance "antenna". So, I question resonance effects
based on a form of a monopole or dipole model, i.e.,
a voltage model. In fact, since the guard traces
are grounded (shorted) at both ends, any "resonance"
in the traces would be theoretically completely
negative in reflection. Or, should be.
So, I don't think the primary mode with which guard
traces operate is with voltage. Of course there's
going to be capacitive coupling. But, with a low
impedance path to ground at both ends, i.e. current
path, IMO this would increase any mutual inductive
effects. As such, the inductive coupling would be
in opposition to the source current causing a
destructive effect of any capacitive coupling. As
vias are placed closer to one another, I see that
as eventually killing the inductive effects.
In most signal traces the dominant coupling is capacitive,
with a secondary inductive coupling. Capacitive coupling
of traces is sensitive to the relative positions of the
traces to each other and to planes. Inductive coupling
stands fairly robust in all cases except separation.
So, in guard traces, the inductive coupling "effect"
simply becomes greater with increased current.
Otherwise, somehow guard traces do something different
than say putting a passive conductive plane between
the plates of a cap without changing the separation of
the plates. As in ...
Cap 1 Cap 2
(+) ----------- (+) -----------
(-) ----------- (-) -----------
Currently, it's the only model I can come up with to
understand how by inserting a passive trace between
two active traces decrease crosstalk even though there's
coupling among all of them.
And that's the way I see the mechanism of guard traces.
So, comments ... ?
Is the model full of it ... ?
What's the reality of this model ... ?
**** 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
This archive was generated by hypermail 2b29 : Wed Nov 22 2000 - 10:50:43 PST