[SI-LIST] What to do with "isolation" conductors in ribbon cable?
- From: "Loyer, Jeff" <jeff.loyer@xxxxxxxxx>
- To: <si-list@xxxxxxxxxxxxx>
- Date: Thu, 27 Sep 2007 13:13:39 -0700
I have a design where I have extra conductors in my ribbon cable
("off-the-shelf, same-thing-we've-been-using FOREVER", no shielding,
ribbon cable), allowing me to put grounds (or whatever I want) between
signals. But, I'm not sure that grounds provide the best isolation.
I'm wondering if those isolation conductors shouldn't be terminated to
ground using 50 ohm resistors (or whatever the impedance of the cable
is).
For example, the pinout may be:
GND S1 ISOLATION_CONDUCTOR S2 GND
Keep in mind that all conductors are identical while they're in the
ribbon cable environment. GND, S1, ISOLATION_CONDUCTOR, etc. have no
idea whether they're supposed to a signal, ground, or whatever.
Is the "ISOLATION_CONDUCTOR" (between S1 and S2) best tied directly to
GND at both ends of the connector, or better tied to GND through a
resistor matching the characteristic impedance of the cable?
If it's tied directly to ground, it seems to represent a resonant
structure. Or, in the time domain, the FEXT/NEXT induced on it would be
coupled to the other signal, and also bounce back at the interface with
ground (rho = -1). But, if half the return current is flowing in it
(which, by symmetry, it is), I don't want the discontinuity at the
resistor.
Terminated at both ends, I think it would couple less energy between the
signals.
The only empirical data I have remotely tied to this is when I had a GND
conductor which was only tied to GND on one end (the other was
floating). Crosstalk was HUGE. Crosstalk was reduced to almost nil
when I tied both ends to GND, which makes me believe the termination
resistors on each end aren't necessary.
All sorts of signals may be going across the ribbon - USB, resets,
clocks, video, etc., but not super fast (though the ribbon may be
significantly long compared to the risetimes).
Any experience with this sort of thing? I'm wondering if I'm
overlooking some basic concept that would make a detailed analysis
unnecessary to predict the outcome. (or you could call it being lazy...)
Thanks in advance,
Jeff Loyer
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
