From: Dave Heald (firstname.lastname@example.org)
Date: Tue Apr 03 2001 - 01:17:31 PDT
First, strip the isulation on the cable back and tie the shield itself
directly to your connector. Drain wires should never be used as an
shield termination (too much series impedance at high frequencies -
these days you WILL have EMC issues). This can be tricky with foil
shields, but you should be able to use something called a "z-fold" (I
think that's what it's called) to get a full perimiter connection to
your connector - talk to your cable manufacturer. For braid or double
shields, it is a lot easier.
OK, now to answer the question. Tie both ends to the chassis. From an
EMC point of view, this extends the functional faraday cage to include
the entire system. Floating one end is not a good idea (acts like a
monopole antenna) but both could be floated only if used with an
appropriate common mode choke and good board layout at the connector
(and you don't need a shielded cable at this point - an example is
I think the single ended idea probably came from DS3 coax. This used
to be required to have a "floating side" in the EU (maybe here too) to
prevent ground loop/ground potential difference problems. The EU has
since started to move to a common bonded network architecture and my
understanding is that both ends can and should now be grounded.
> If you are running differential signals between to racks through a
> shielded cable, what should you do with the cable drains?
> * Hard wire both-ends to the chassis?
> * Hard wire one end and let the other float?
> * Float both ends?
> * Connect through a series resistor, or series cap, or series cap
> and resistor in parallel?
> Is there a rule of thumb or formula based on something to help
> you determine how to connect the shielding?
> I have seen it done several different ways but never have been
> real sure which way is correct.
**** To unsubscribe from si-list or si-list-digest: send e-mail to
email@example.com. 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 : Thu Jun 21 2001 - 10:11:24 PDT