Dear Ted Your description of ground-loop prevention methods is very comprehensive, and reminds me strongly of the years I spent designing consoles for Neve in the 70's - 80's. Thank you very much for sharing this useful information with us.
These days most of my involvement with pro-audio is helping manufacturers and contractors to design and build equipment, systems, and installations that meets the EMC Directive and its product-family standards EN 55103-1 (emissions) and EN 55103-2 (immunity). What we have found over the last five-six years is that the shields of audio cables need to be RF-bonded at both ends, otherwise they become resonant at cable lengths of quarter-wavelength and longer (e.g. cables of around 0.75 metres or longer at 100MHz) and cease to provide any shielding. So many pro-audio products now have digital processors in them, and common-mode clock and data bus harmonics can leak out of inadequately-shielded audio cables, causing problems for emissions compliance. All audio opamps will demodulate happily up to several hundred MHz (bipolars are worse than FET types) and as a result fail the continuous RF immunity tests if they have inadequately shielded cables. Of course you can fix these problems with filters instead of shielding, but it is more cost-effective to fit smaller filters and make the cable shielding that is already in place work better at RF. I have worked with a number of highly respected pro-audio console and outboard equipment companies on this problem, and some of them are now able to supply legitimately CE marked equipment and systems of any size, with absolutely the highest quality sound, with great ease of installation - with all their audio cable screens electrically bonded at both ends and ground loop currents flowing in them without creating any hum problems whatsoever. Sounds good, eh? All the very best! Keith Armstrong www.cherryclough.com In a message dated 21/03/02 00:01:32 GMT Standard Time, [email protected] writes: > Subj:GROUND LIFTS, oh boy..... > Date:21/03/02 00:01:32 GMT Standard Time > From: [email protected] (Ted Rook) > Sender: [email protected] > Reply-to: <A HREF="mailto:[email protected]";>[email protected]</A> (Ted > Rook) > > > > > X-INFO: INVALID TO LINE > > It is unfortunately the case that it is easy to create a sound system > having mediocre performance using ready-made audio cables with shields > connected at both ends. > The technical knowledge required to install a system that performs > correctly and is safe is available, at a cost, and that is where the risk > taking behavior with protective grounds enters the picture. > Untrained people try quick and dirty methods to solve what can be genuinely > complex electrical installation requirements. > Between microphone and power amplifier there is commonly 1000x voltage > gain. Plus another 30x to 100x in the power amplifier. So 1mV noise on the > signal input common would become several volts at the speaker, deafening. > For good performance the signal input common must be clean of hum and RF > down to microvolt levels, yes micro-volts. And you computer engineers think > you have problems with voltage discrimination on 3V logic....? > > seems like some education is in order, ground loops 101: > > > The problem with ground loop hum pickup is real and the solution requires > that each piece of equipment is grounded once, and once only. > Ground loops get created by inexperienced installers when they use audio > cables with the shields on at both ends to hook up AC powered gear that has > chassis and signal common connected. > > If you ever installed a room full of audio gear, or a building full, you > know that there are several techniques for controlling hum pickup and one > is to eliminate ground loops. > A ground loop is created when there are two paths for ground between two > physically separated pieces of equipment. > The most common loop is from the AC ground at the first outlet, up the line > cord onto the chassis. > Through the chassis onto signal ground. > Out of the first unit via the signal shield. > Along the signal shield to the signal ground of the second item. > From signal ground to chassis in the second unit and from chassis down the > AC line cord to the AC ground of the second outlet. > This path is supposedly at one potential throughout but that is impossible > in the presence of a 50/60Hz AC field due to the magnetic field radiated > from AC powered circuits. > Recall school physics? > When a conductor cuts magnetic lines of force an emf is generated. > The physical size of the loop and the strength of the 50/60Hz field in the > vicinity will determine how much current flows in the ground loop. > If this dirty ground is allowed to become the signal common at the input to > a device with gain it will appear at the output amplified. > In the case of multichannel devices like recorders and mixers the hum > signal will be coherent at all inputs and will sum at the output. > > The loop can be broken in many places. > > The sensible place to break it is by lifting the connection between signal > ground and chassis in either unit or by not connecting the signal shield at > both ends of the audio cable. > > When all the equipment has balanced audio ins and outs there is little > difficulty because you can separate signal common from chassis. > > Problems arise with lower cost unbalanced items which have signal common > and chassis combined. > > The correct technique with these is to introduce audio isolation > transformers on the signal lines, not cheap, not always easy. > Alternatively a rigidly enforced single point grounding scheme for all > protective grounds and signal commons will also work. > This requires the preparation of a plan, drawings, skilled installation > labor and supervised maintenance. > No problem so long as you have the budget. > > One quick and dirty way to break loops as a temporary measure with > unbalanced gear is to allow protective ground to be connected via the audio > shield to the next item's chassis. > Then the AC third wire on the unbalanced item is lifted to break the loop. > This has a serious vulnerability. when the audio is unplugged the > protective ground may be interrupted. > For this reason you will only find the quick dirty fix in unsupervised > installations such as home studios and small time theater. > The technician who does this stuff is usually untrained and working on an > unpaid basis doing the best he can. > The alternative, to design and create single point grounded wiring for the > complete audio installation, is beyond the resources of such facilities. So > they live with the risk. Note that if this occurs it is because the > installer overrides the manufacturers instruction to connect a protective > ground. > In my experience no one goes to the trouble of opening the equipment to > disconnect the protective ground, it is done at the AC outlet by defeating > the protective ground connection. > Anyone who has ever done this knows how uncomfortable it feels and how it > promotes efforts to fix the problem with one of the proper safe solutions > as a matter of urgency. > > .....I find the claims of some posters today about defeat-hardware they > have seen for protective grounds to be unbelievable irrational and > inflammatory. > In my opinion you will not see pro audio equipment with a method of > defeating the protective ground. It would be unsafe. > I challenge people who say they have seen such things to give the > manufacturer and model of the hardware. > What you may have seen was a signal ground defeat method not a protective > ground defeat method. > > Would you want to expose your sound system to the risk of going live? No, > nor does any amateur or professional sound system installer or operator. > > Someone posted that they believed hum problems could be solved by 'proper' > grounding. > The implication is that sound installers are failing to install adequate > earthing conductors and using defective plugs. > Apart from being disrespectful this overlooks the reality of extended > ground loops and their interference effect on signal common. > There are several reference sources you may prefer to trust on the subject, > the AES.org website may be a good place to begin for sound studio practice. > PA sound systems for fixed installation and for touring require application > of the same practice often on a bigger scale. > In the US the NSCA may be able to provide some verification of what I have > written above. > > Every public venue, sports arena, recording studio, radio broadcast station > and theater in every country on the planet has had to deal with the > realities of installation for high quality audio distribution but > unfortunately the skills and knowledge are still not widely known. > > > Best Regards > > Ted Rook, Console Engineering, ext 4659 > > Please note our new location and phone numbers: > > Crest Audio Inc, 16-00 Pollitt Drive > Fair Lawn, NJ 07410 USA > > 201 475 4600 telephone receptionist, 8.30 - 5 pm EST. > 201 475 4659 direct line w/voice mail, 24 hrs. > 201 475 4677 fax, 24 hrs. >
