Telco equip is capable of this problem in many instances. Once at a former company, we had two power outlets each from a different substation in the lab. We didn't realize that a float of more than 15VAC existed between them until we setup a system and actually saw the coax between them melt. That was the result of improper connection of ground by the electricians in each of the substations inot the facility. We fixed that problem by capacitvely coupling one end of the shield to the shell.
Years later, in the telco world, a similar complaint came up from a customer. When faced with a massive redesign of many systems, I resorted to the "we are not responsible for the poor wiring management of your facility ..." and convinced the customer that they had the potential of a major problem not indigenous to just the equipment of concern. They agreed. To this day I have no idea if I did the right thing. It is my "gut feeling" that ground loops at the customer site is of a major concern for the customer and should be resolved by the customer. Ground loops should of course be avoided with any equip, but that's just one side of the story. Doug On Fri, 6 Nov 1998, Terry J. Meck wrote: > This brings up a question I have been fighting with ever since I got > involved with EMC compliance testing. > > > A product is designed and complies with a 360 degree shielded cable. > An interrogator buys a group of products and goes to install them in > a large facility. Each product is specified to require a shielded, > let's say, communications cable RS 422 or 485 etc. But they may be > separated by 100s of feet. Obviously not connected to the same > power legs. How does the integrated prevent the ground loops and > still maintain EMC compliance? > > He is buying CE, FCC compliant products. No short cuts taken but I > can envision anything from poor signal quality to smoke in the > installation with the potential ground loops I envision. > > Maybe you can give me fewer sleepless nights worrying about these > poor integrators. :-) :-) > > > On 5 Nov 98 at 22:20, Cortland Richmond wrote: > > > Rick, > > > > I don't see that you have a problem. Your differential video > > signals are > > not susceptible to ground loops, indeed,they do not require ground > > at all. You can safely rely on the shield to function only as a > > shield, and not as a signal return. This makes your job easier. So > > I would answer your questions as follows: > > > > 1. No, in this case it is not reasonable to terminate only one end > > of the shield -- as you have discovered yourself, doing so increases > > radiated EMI. 2. No, differential returns can not be successfully > > grounded to chassis return -- that would upset the balance -- and > > should not be. 3. Given the distance between your devices, you may > > have to use capacitive decoupling on the shield at one end. There > > are coaxial connectors made which include planar, low-inductance > > capacitors between the shell and actual ground. This might do for > > your application. 4. I always wanted to terminate the signal > > returns at the device which generated the signal, so the return > > currents flowed only in the signal return wire and not in the > > chassis. This seemed to reduce common-mode problems. > > > > Regards, > > > > Cortland > > > > > > > > ====================== Original Message Follows ==================== > > > > >> Date: 04-Nov-98 11:13:23 MsgID: 1065-13982 ToID: 72146,373 > > From: [email protected] >INTERNET:[email protected] Subj: Cable > > Shielding Chrg: $0.00 Imp: Norm Sens: Std Receipt: No > > Parts: 1 > > > > > > My company manufacturers image generation equipment (large mainframe > > computer) which is typically placed at a significant distance from > > the display device. These may be as much as 50 to 150 feet apart. > > In my EMC tests, I have determined that due to common mode, harmonic > > noise, both ends of this shielded cable must be terminated to the > > respective device chassis. Unfortunately, this has the potential to > > create significant ground loops which severely compromise the > > quality of the video signal. To add insult to injury, the signals > > within the cable are twisted pairs, differentially driven. This > > eliminates (as I understand it) the possibility of terminating the > > signal return to chassis ground at the point at which the cable > > exits the enclosure. So, on the surface it appears that I have the > > choice of meeting EMC regulations or meeting equipment performance. > > > > > > My questions are as follows; > > > > 1. Is it reasonable to provide single end termination of the overall > > (outside) cable shield? 2. Can differentially driven signal grounds > > be successfully grounded to chassis return? Should they be? 3. Is it > > reasonable to request or demand that both devices are sourced from > > the same power AND/OR bonded to a common power grid? 4. Is it > > desirable for signal returns to terminate at the chassis ground at > > the point at which the cable exits the device. > > > > > > Thanks in advance > > > > Rick Busche > > Evans & Sutherland > > [email protected] > > > > --------- > > > > --------- > > This message is coming from the emc-pstc discussion list. > > To cancel your subscription, send mail to [email protected] > > with the single line: "unsubscribe emc-pstc" (without the > > quotes). For help, send mail to [email protected], > > [email protected], [email protected], or > > [email protected] (the list administrators). > > > Best regards, > Terry J. Meck > Senior Compliance/Test Engineer > Phone:215-721-5280 > Fax:215-721-5551 hard copy; > Fax PC: 215.799.1650 To my desk PC > [email protected] > Accu-Sort Systems Inc. > 511 School House Rd. > Telford, PA 18969-1196 USA > > --------- > This message is coming from the emc-pstc discussion list. > To cancel your subscription, send mail to [email protected] > with the single line: "unsubscribe emc-pstc" (without the > quotes). For help, send mail to [email protected], > [email protected], [email protected], or > [email protected] (the list administrators). > > --------- This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to [email protected] with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to [email protected], [email protected], [email protected], or [email protected] (the list administrators).
