Hi Arlen, What you've done is a wonderful thing. Ground loops are NOT a worry in the digital, high frequency world; they're a red herring. Ground loops ARE a problem if you have low frequency instrumentation, then it is often necessary to isolate earth ground connections on shields and so forth. This is primarily done to reduce circulating 60 Hz currents in instrumentation sensor cabling.
Digital signals have a high amount of inherent noise immunity to any 60 Hz that might be floating around. We recommend your approach when fixing, designing and troubleshooting equipment. A common point grounding approach is difficult (impossible!) to implement when dealing with high frequency circuits and systems. Multi-point grounding is the way to go, with low impedance connections to minimize RF potentials. Best regards, Mike Violette Washington Laboratories, Ltd RF, EMC and Safety Testing (301) 417-0220 [email protected] WEB: http://www.wll.com ---------- > From: Arlen Olive <[email protected]> > To: [email protected] > Cc: arleno (E-mail) <[email protected]> > Subject: Grounding vs. Ground Loops > Date: Thursday, December 11, 1997 2:17 PM > > I've been working on a multimedia board in a standalone box > that connects to a PC parallel port. It also has an external > DC power supply, and connections for audio and video > input and output. > > To pass FCC class B, I had to: > 1. Shield the box with a conductive coating connected to the board ground > 2. Connect the board ground through the DC power supply's DC return to Earth Ground > 3. Add a substantial ground connection between the board ground and > the parallel port connector shell. > > Some of my colleagues have expressed concern because this creates a ground loop > between the PC (through the parallel port shield) and the DC power supply Earth Ground, > not to mention potential ground loops to the audio/video input/output devices. > > One suggestion has been to use AC paths (i.e. capacitors) to the board ground for > everything except the power supply return. I have serious reservations about this approach, > (I'm convinced it won't pass FCC) but I don't really know what to say about the > ground-loop argument. I'm well aware of the importance of common-point grounding, > but I haven't found any other way to bring the EMI emissions down to class B levels. > Any suggestions? > > (Directions to any good reference material would be greatly appreciated.) > > > Arlen Olive > Hardware Engineer > FutureTel, Inc. > 1092 East Arques Avenue > Sunnyvale, CA 94086-4601 > Phone: 408-522-1472 > FAX: 408-522-1439 >
