>From the I remember when file...... In the olden days - before integrated magnetics at least - the Ethernet ground was moated with a small gap between that and the Ethernet output components, outgoing circuit traces, connector etc. The PCB area after the magnetics was a keep out area for power and ground. The RJ45 could be grounded the chassis, I think, but the PCB's digital ground had to be kept on the other side of the moat. The capacitance to ground figures for data wires is in the 6 nFd ranged if I remember right, and that pretty much eliminated a lot of filtering on the output. The moating prevented a dirty ground from being coupled into the Ethernet lines. Theoretically at least the Ethernet lines didn't have shielded although and I think the specification called out UTP - unshielded twisted pair. I've been in a lot of discussion whether or not UTP is the same as over cable shield - foil or braid - so somebody was trying to shield the line, and you can buy RF-45's that have shielding and those that do n! ot. The magnetics crossed the gap and were center tapped - and hopefully balanced. Each side of the outgoing half of the transformer had filter caps if those caps didn't match very well then there was a significant amount of CM noise generated and the emission got pretty bad. That was in the early days of Ethernet and things change as they improve - like integrated magnetics so I'm much less familiar with those problems, and separated in time long enough that I can't even remember the whole suggested Ethernet layout. As a differential signal the drive signal PCB trace routings had to be matched and the board impedance controlled. I remember a great application note that Neven (you still out there?) wrote for Intel years ago. Using cheap components did help anything either.
Its been awhile since I've been involved so there may be some small errors in the above and as I said as things (other than me) age they get better and more solutions to preventing the original problems are developed. -----Original Message----- From: Oglesbee, Robert A [mailto:[email protected]] Sent: Wednesday, September 29, 2010 1:29 PM To: [email protected] Subject: Re: [PSES] EMI issue with Ethernet How the transformer is wound also makes a big difference. For integrated magnetics, changing jack vendors can really affect the spectrum. Although the difference is not dependable... Regards, Rob -----Original Message----- From: John Woodgate [mailto:[email protected]] Sent: Wednesday, September 29, 2010 4:04 PM To: [email protected] Subject: Re: [PSES] EMI issue with Ethernet In message <[email protected]>, dated Wed, 29 Sep 2010, "ce-test, qualified testing bv - Gert Gremmen" <[email protected]> writes: >Most of the time the problem is that the ground point centertap > >of the Ethernet transformer is insufficiently grounded. This may well be a practical finding, but ground really has no function for a balanced transmission line, which is what Ethernet (and several other techniques) is based on. Ground at one end is never at the same potential as ground at the other end, especially at RF. - ---------------------------------------------------------------- This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list. To post a message to the list, send your e-mail to <[email protected]> All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc Graphics (in well-used formats), large files, etc. can be posted to that URL. Website: http://www.ieee-pses.org/ Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas <[email protected]> Mike Cantwell <[email protected]> For policy questions, send mail to: Jim Bacher: <[email protected]> David Heald: <[email protected]> - ---------------------------------------------------------------- This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list. To post a message to the list, send your e-mail to <[email protected]> All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc Graphics (in well-used formats), large files, etc. can be posted to that URL. Website: http://www.ieee-pses.org/ Instructions: http://listserv.ieee.org/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Scott Douglas <[email protected]> Mike Cantwell <[email protected]> For policy questions, send mail to: Jim Bacher: <[email protected]> David Heald: <[email protected]>
