HI All, thanks for the great suggestions. I didn't get chance to work on this yesterday too much, but hope to this evening ( USA time ).
Just to clarify, this is a small plastic box about the size of a matchbox with 3 connections. Ethernet, RS 232, and low voltage dc from a wall wart. I did observe that putting a ferrite bead on each cable made little ( maybe 1 dB ) difference. Nor did wrapping the box in AL foil and bonding the foil to the shield. The emission is a single spike @ 125 MHz about 20 to 25 dB higher than surrounding noise. The cables are brand new, and have foil and braid. I hope to share the findings tomorrow, Sincerely, Derek Walton [email protected] wrote on 10/5/2006, 7:39 AM: > Hi Derek, > > I've seen a couple of posts with good advice. I'm wondering if you've > already taken care of bonding the shield; and you're still having issues. > > Forgive me for repeating your question...but I'm thinking out loud (which > is difficult for email :-)...for some reason when you just have the > Ethernet cable plugged into the transmit end (The DUT) in the chamber... > the Ethernet signal conductors don't change potential at 125Mhz with > enough magnitude to couple to the measurement antenna...but when you > terminate the receive end, outside the chamber..they do. > > That could be due to the symbol rate of the actual ethernet > traffic...which would probably increase when the receive end is plugged > in. Could it also be a coincidence...how good does the receive end > filter > its Ethernet cable? Could the receive end have another, unrelated clock > at 12.5 or 25Mhz that's catching a ride out? > > Now...somehow, the emissions get higher with a shielded cable...is this > with or without the receive end plugged in? > > For the shield to work best...it would need to be bonded at both ends to > the same, infinite capacity, zero resistance ground at Earth > potential...which would also, hopefully, be the same ground as the shield > of the coax for the measurement antenna...this would minimize measured > emissions. > > Any break in that chain...and you have an issue. I saw "ground currents" > mentioned. This would be due to some potential difference between the > shield's end at the DUT, the shield's end at the receiver and or the > ground of the measurement antenna system. > > If you've already messed with shield bonding...Have you also tried the > following experiments? > > -Are you sure of the shield integrity within the cable? Is it old? > Has it > been bent a few times? > -What if you try a different device at the receive end? > -Have you played with the grounding of the chassis of the device at the > receive end? (grounded, ungrounded...grounded back to chamber ground) > -What if you leave the cable hooked up at the receive end, leave it > running into the chamber and disconnect it from the DUT? What happens? > -Are there any power filters that could possibly be breaking the Earth > path between the DUT and the device at the receive end of the Ethernet > cable? > > ---What if you just turn it all off, go to the local pub and catch a ball > game over some food and spirits...maybe the emissions will still be > there; > but you won't care :-) > > I don't know what's at the receive end; or what your DUT is; but many > Ethernet devices are powered by Wall Warts and carry no real chassis > ground. Many "shielded" ethernet cables have wimpy shielding and poor > termination to the connector...to the point where the termination to the > connector could be ruined after some hard bending cycles. Running this > type of shielded cable between two devices with unknown ground > characteristics is always "fun". I don't know if I've ever gotten a > device through a suite of EMC testing unscathed if it had an Ethernet > cable...seems as though we always had to fiddle with it; either at our > end, the peripheral end, shielded, unshielded... very similar to the > issues that you're having now...I just don't remember the differences > being so pronounced as what you're seeing. > > The other problem with Ethernet is that it's a high frequency signal, so > the ports are usually poorly filtered (for EMC anyway), which means that: > if somebody didn't take care of business with other clocks in the system, > they can easily catch a ride out on the Ethernet cable. That's why I was > wondering if the 125Mhz measurement is only coincidentally related to an > Ethernet frequency. > > Hope this helps and good luck. > > Chris Maxwell > Electrical Engineer > Critical Imaging LLC > > > - This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list. Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to [email protected] 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] All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________
