From: [email protected]
Date: Mon Nov 27 2000 - 06:44:51 PST
Ellis,
We have done a number of 10BASE-T/100BASE-TX designs in the last four years.
Several items that we have found to be very critical in layout are:
* Minimize the distance from the PHYceiver to the connector-- we have found
that just 1/2 inch increase in the length of
the traces from the connector (via common-mode-choke and transformer/filter)
to the PHYceiver can reduce the
maximum length of cable we work with by 30%.
* Keep *everything* away from the Ethernet signals-- cut a hole, in every
layer, that extends from the signal pins of the
RJ-45 connector to halfway between the rows of pins of the
transformer/filter. Make this hole as wide as the
transformer/filter and common-mode choke. Do not permit anything except the
Ethernet components and traces in
this area.
* Try to have only those components and traces directly associated with the
Ethernet signals between the
transformer/filter and PHYceiver, although we do permit power and ground
planes in this area.
* Run the Ethernet traces parallel from the connector to the
transformer/filter. If they must be twisted to match the
pinouts, do so between the transformer/filter and the PHYceiver.
For Radiated Emissions, we design our boards with a four-winding common-mode
choke between the transformer/filter. Because of their cost, we prefer not to
use them if we don't have to. So, for our first-pass boards, we either:
* Short the windings with traces on the topside of the card-- which can be
cut, then a common-mode choke populated if
needed.
OR
* Place 0603 0-ohm topside resistors in parallel with the windings--- which can
be removed, and the common-mode
choke populated if needed.
John Barnes Advisory Engineer
Lexmark International
**** 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 : Tue May 08 2001 - 14:30:15 PDT