Hi Chris, I'd just like to add a few words since I have experience which is counter to some of the opinions posted.
First, if shields didn't work, we would never used them... Right? Second, the effectiveness of a shield depends on several factors including the nature of the noise source, the nature of the victim and the shield chosen. Third, employing some shield topologies could introduce undesirable side effects like ground loops. This is really a problem for Audio, and not really an Ethernet issue unless you go long distances. If you do, then go optical. Fourth, incorrect opinions are formed because of some unknown action when something was tried. Fifth, too much credence is given to what SHOULD happen in a perfect world... So, here are my specific comments on Ethernet in a noisy environment: We have tested many insitu machines with Ethernet. Without an exception the machines having shielded cables had less noise issues. We have, in the lab, tested many vendors Ethernet products. A significant number performed better with shielded cables. This included routers from Cisco, Belden and Netgear that came in as support equipment. The two exceptions were Ethernet to RS-232 and RS-485 converters. For my main client when testing his products, shielded Ethernet rarely caused issues, even performing EFT testing at 4 kV. Issues observed were almost always the PC locking up though noise getting in via the keyboard or mouse cables. In the lab we used to have an unshielded network. This would trash every time EFT was performed. We now use a good quality Shielded Ethernet cable and have not had one event in 4 years. I have not seen issues with ground loops... Sincerely, Derek Walton L F Research -----Original Message----- From: Chris Wells <radioactive55...@comcast.net> To: EMC-PSTC@LISTSERV.IEEE.ORG Cc: christopherdwe...@eaton.com Sent: Sun, Nov 28, 2010 7:48 pm Subject: RE: [PSES] Shield bonding on STP Cat5 Ethernet cables Thanks all I've received some good feedback on this topic. Seems that most of you don't see a lot of value in SHIELDED - STP Cat 5 cable over well balanced UTP. I understand the value of a balanced cable and how it leads to a solid CMRR. At the same time I have seen shielded cable make a difference with those RS485 applications so I am not letting go yet. The power distribution gear I am talking about will include Motor Control Centers with Adjustable Speed Drives and Soft starts. Also Lighting supply panels feeding electronic Ballasts and a variety of interesting loads found in Commercial and Industrial applications. All sorts of continuous diode commutations driving current pulses and then a variety of surge and fault events that happen less frequently. We typically test our equipment to level 3-4 for the standard array of IEC61000-4-X tests. I did point out that we presently try to use STP only in locally well bonded power distribution line ups. I would not expect that there would be any significant ground loops between local nodes. Under those restrictions do you still think STP is not useful? The assembly gear has little room for special Panduit or conduit chasses for specialized Ethernet communications. Our RS485 serial communication cables are typically strapped to control cables with ratings up to 600V. Belden cable company has developed some 600V AWM STP cables which have either a Beldfoil shield or a combination Beldfoil and 70% TC braid shield on top of their very balanced cable. AWM stands for Appliance Wire by Manufacture and is basically a UL recognition that needs to be evaluated in the assembly gear for approval. This would allow us to strap the STP Cat5 cable to the control bundles like our RS485 applications. I do not know if the STP nature of the cable helps in achieving the 600V AWM rating or not. It may be possible to achieve this rating without the shield, I will check. Chris Wells Eaton Corp. -----Original Message----- From: emc-p...@ieee.org [mailto:emc-p...@ieee.org <mailto:emc-p...@ieee.org?> ] On Behalf Of o. laney Sent: Thursday, November 25, 2010 12:34 PM To: j...@jmwa.demon.co.uk Cc: EMC-PSTC@LISTSERV.IEEE.ORG Subject: Re: [PSES] Shield bonding on STP Cat5 Ethernet cables A magnetic field will couple through the shield even if grounded at both ends, albeit in attenuated form. It's really just a matter of how much attenuation can be achieved. I've been working on an instrument with a pulsed magnetic field powerful enough to cause robust electrostimulation of any flesh in the near vicinity. An interposed sheet of ordinary kitchen grade aluminum foil reduces but does not eliminate the shock sensation. This is at 5 KHz. Turning to the matter of audio cables, hum levels even 60 to 80 dB below the program content will prove annoying, and the standard braided shield or even braid over foil can reduce the hum, but at power frequencies the answer is never as much as one would wish for. Absent resort to mu-metal shielding, the key to audio hum rejection is, as always, balance rather than shielding per se. The physics is straightforward enough. In magnetically transparent metals such as aluminum or copper, magnetic shielding is achieved by eddy current effect rather than flux shunting. Most shields are rather too thin to be efficient at this for power frequencies. For instance, the skin depth for pure copper at 60 Hz is around 8.5mm. It takes 4 skin depths to reduce ambient magnetic fields by 70 dB, easy at RF but rather impractical for hum reduction in an audio cable routed past a power transformer, or control wiring routed near 'cabling carrying large currents'. I completely agree that UTP is satisfactory for ethernet even in noisy industrial environments. It's not just that the system is well balanced, but also that the signals are transformer coupled and galvanically isolated from the equipment with insulation good past a kilovolt. Adding a shield is just a means to violate galvanic isolation. At RF a properly grounded shield can help suppress CM radiation, but if that is a problem then the balance assumption has been blown and ferrites are probably a better solution. Why are shielded CAT-5 cables available? Because people are willing to buy them, and to keep EMC consultants busy when it doesn't help. Orin Laney On Wed, 24 Nov 2010 21:34:10 +0000 John Woodgate <j...@jmwa.demon.co.uk> writes: > In message > <FCA549BE3ECF9D4CB8CB8576837EA4890A7744@ZEUS.cetest.local>, > dated Wed, 24 Nov 2010, "ce-test, qualified testing bv - Gert > Gremmen" > <g.grem...@cetest.nl> writes: > > >Magnetic fields from will couple > >through the shield. > > Well, not if it's grounded at all frequencies at both ends, but then > you > get all the circulating currents problems. > > We all seem to agree - use UTP unless you find you can't, but using > STP > may be difficult anyway. > -- > OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and > www.isce.org.uk > John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK > If at first you don't succeed, delegate. > But I support unbloated email http://www.asciiribbon.org/ > > - > ---------------------------------------------------------------- > 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 <emc-p...@ieee.org> > > All emc-pstc postings are archived and searchable on the web at: > http://product-compliance.oc.ieee.org/ > 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 <emcp...@radiusnorth.net> > Mike Cantwell <mcantw...@ieee.org> > > For policy questions, send mail to: > Jim Bacher: <j.bac...@ieee.org> > David Heald: <dhe...@gmail.com> > > - ---------------------------------------------------------------- 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 <emc-p...@ieee.org> All emc-pstc postings are archived and searchable on the web at: http://product-compliance.oc.ieee.org/ 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 <emcp...@radiusnorth.net> Mike Cantwell <mcantw...@ieee.org> For policy questions, send mail to: Jim Bacher: <j.bac...@ieee.org> David Heald: <dhe...@gmail.com> - ---------------------------------------------------------------- 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 <emc-p...@ieee.org> All emc-pstc postings are archived and searchable on the web at: http://product-compliance.oc.ieee.org/ 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 <emcp...@radiusnorth.net> Mike Cantwell <mcantw...@ieee.org> For policy questions, send mail to: Jim Bacher: <j.bac...@ieee.org> David Heald: <dhe...@gmail.com> - ---------------------------------------------------------------- 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 <emc-p...@ieee.org> All emc-pstc postings are archived and searchable on the web at http://product-compliance.oc.ieee.org/ 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 <emcp...@radiusnorth.net> Mike Cantwell <mcantw...@ieee.org> For policy questions, send mail to: Jim Bacher <j.bac...@ieee.org> David Heald <dhe...@gmail.com>
