Associated Equipment:
Front End
Digital Front End
Power Conditioning

A Well-Grounded Education


Tom Mallin

November 2003

I'd like to suggest that in order for the performance of your system to soar to new heights, you must first pay close attention to the way you tether it to the earth below.

Electrical code provisions and textbook advice on proper grounding usually deal primarily with life and property safety issues, as well they should, rather than with audible consequences. If sound is mentioned at all, it is usually in the context of best practices for removing ground loop hum at 60 Hz and its harmonics, with little or no mention of potential affects of any particular grounding scheme on other aspects of reproduced sound quality.

Building power circuits using 110/120-volt single-phase alternating current and high-quality hospital-grade outlets ideally contemplate using three wires, two of which are supposed to be at ground potential: the signal ground and the earth ground wires. Some electronic equipment comes with two-prong connections which lack any sort of built-in ability to connect to the earth ground connection of the building's power outlet. Others have three-prong connections providing a means of connecting the unit via its power cord to both the signal ground and earth ground connections at the building's power outlet.

Most experts agree that there should be a single earth-ground reference point for a building's electrical service. Most recommend some sort of copper rod or pipe embedded in moist soil near the building's electrical service entrance. This copper rod or pipe should stick up out of the ground a few inches and be connected by heavy copper wire at that point to the ground terminal of the electrical service box. Most experts would also agree that all the signal and earth ground wires for the various circuits in the building should be carried on separate wires and star-grounded back to this ground terminal with copper wire of at least 12 gauge. I have done all this and in fact am able to establish a direct copper-wire-to-copper-wire connection between all my 10-gauge copper signal and earth ground wires and the huge-gauge ground terminal wire in the service box—they are all clamped together in the service box at a single clamping junction.

Where the experts begin to equivocate is when they discuss what to do about electronic components which have no earth-ground prong on their power cords, or what to do if multiple components in an audio system have three-prong power cables. Should all these power cords be connected straight into electrical outlets? Should some or all of the three-prong plugs be "cheated" so as to interrupt the earth-ground connections to the wall outlets and, if so, what is the proper component(s) to leave connected to earth ground? Should some or all equipment lacking three-prong power cables be earth grounded to the electrical outlet's earth ground connection with separate wires running from electronic chassis grounding points to the earth-grounding connection of the outlet or to some other earth ground like a cold water pipe?

The choices are further complicated once so-called "balanced power" schemes (not to be confused with balanced interconnections between components using XLR-terminations) are introduced. The PS Audio Power Plants are probably the most popular of these among American audiophiles, but such units are also produced by API and Equi=Tech, among others. Martin Glasband, the head of Equi=Tech, is probably someone to be reckoned with in this field, with over 30 years of hands-on electrical engineering design work with commercial sound studios. He was also instrumental in getting the U.S. national electrical code revised to condone "balanced power" wiring schemes. Some of Glasband's comments on grounding are at


Basically, Glasband states that proper grounding of systems using "balanced power" devices like the PS Audio Power Plants I use entails earth grounding ALL components. However, even Glasband's comments don't specifically deal with proper grounding of a system which, like mine, uses balanced power sources for the front-end electronics and traditional wall-outlet power for the amplifiers.

My understanding of "classic" grounding theory (that is, pre-"balanced power"), as well as my hands-on experience, indicates that for best sound an audio system usually should have a low-impedance connection to a true earth ground at only one single point, with all the other components in the system deriving their earth-ground reference through the signal grounds of the interconnecting cables. In applying this theory, I have treated power regenerators like the two PS Audio Power Plants I use for my front-end electronics as mere extensions of the wall outlets. I thus fully earth ground the Power Plants, and have usually cheated the earth grounds of components plugged into the Power Plants. I also earth ground my MIT Z-1 Stabilizer parallel line conditioner since it is not physically or electrically directly connected to any other electronic component in the system; I also earth ground it because it is my understanding and experience that part of the device's good effects come from its operation on the earth ground connection and that it must be directly connected to earth ground to do this.

I had also determined through trial-and-error experiments over the years that the amplifier is sonically the best component to connect to earth ground, not the preamp (as is stated in some texts) or any other component of the system. While one-time TAS writer, Enid Lumley, is credited by Laura Dearborn's classic audiophile text, Good Sound (William Morrow & Co., Inc., New York, NY, 1987, p. 229) for the amplifier-is-the-best-spot theory, this is an error. Enid Lumley was in fact adamant in her writings that ALL earth grounds should be cheated for best sound. And while I agreed with her in 1987, once I moved to a residence which I had made sure had a properly implemented copper grounding rod as an earth-ground reference, and as the environment has become increasingly polluted with RFI and EMI from modern electric and electronic devices, classic single-point grounding at the amp had always sonically trounced floating all the earth grounds.

Theory is one thing. Application to a particular system is another. Description of the audible effects is yet another step.

Let's take my relatively complex system as an example. My recent replacement of the four Bryston 7B-ST amps which drove my Legacy Audio Whispers in bi-amped, bi-interconnected mode with the new 7B-SST model presented something of a grounding challenge. Grounding my system for best sound with the 7B-STs had been a snap. I had simply used the ground lift switches on those amps to lift the earth grounds of the power cord connections on three of the four amps, leaving the right channel bass amp, and the entire system, fully earth grounded at only that point.

Recognize that, in the 7B-ST amps, the ground lift switches did not actually disconnect or cheat the earth ground connection to the amp. Rather, the switch placed a small resistor in series with the earth ground connection, creating a high enough impedance to ground to prevent ground-loop hum, but not so high as to create any problems for life, health, or property safety.

One of the changes Bryston made in the SST was to incorporate a "proprietary grounding scheme," eliminating the ground lift switches. Internal chassis inspection lead me to conclude that perhaps the "proprietary grounding scheme" Bryston had adopted with the SST series was nothing more than a permanent wiring-in of the "ground lift" arrangement which had been available through the switch on the ST series. If this were true, perhaps all I needed to get the best from the SSTs was to somehow duplicate the grounding method I had used with the STs.

To do this, I plugged three of the 7B-SSTs straight into the wall outlets as recommended in the manual. With the fourth amp, the right channel bass amp, I cheated the earth ground connection through the power plug with a 15-amp-rated Eagle three-to-two-prong adaptor.

I then created an extremely low impedance earth ground connection between that fourth amp and the earth ground of the wall outlet. I soldered one end of a four-foot length of 8-gauge hardware store wire to the flat outside of a Cardas RCA cap and pushed the cap onto the ground sleeve of the unused RCA input jack of the amp.

The other end of the wire was connected to a Bryston expanding locking banana plug. This locking banana plug could be expanded just enough to form a fairly solid and tight connection inside the tunnel of the round earth ground connection of the hospital-grade electrical outlet. I thus created a single-point connection of the system's signal ground to an earth-ground reference. Finally, I plugged my MIT Z-1 Stabilizer into the other receptacle of the duplex outlet into which I plugged the new earth ground wire so that the Z-1 could work its magic on both the hot side of the circuits and on the relevant signal and earth grounds.

At first, all seemed rosy. But as the system warmed up and stabilized (at least a four-day process, in my experience) nagging sonic problems arose. The entire bass range seemed to thin out, and the soundstage seemed to develop a bit of a hole-in-the-middle effect. Further, there was added brightness and grit on massed strings and the sound of high brass.

I then eliminated the cheater plug on the right channel bass amp, leaving my special low-impedance ground wire in place. This eliminated the bass thinness, but the high frequency problems remained, less egregious than before because the tonal balance was now better.

More experimentation was clearly called for. As I tried different grounding schemes, I left each grounding change in place for at least a week to make sure the sound of the system was fully stabilized after each power-down/power-up cycle and the cleaning and treating of the changed connections with ProGold GX-5 spray which accompanied each change. Note that none of the grounding schemes I tried produced any audible hum with my system.

I tried cheating the earth grounds of all the SST amps, first without my special low-impedance ground wire in place. Now the upper bass and lower midrange were decidedly too thin and the upper midrange and lower highs were elevated. A very detailed sound, yes, but not realistic. There also seemed to be a bit of gritty distortion in the highs which had not been present with the 7B-STs, as well as a vagueness of image placement which was definitely not there with the old STs. In addition, subjectively the amps seemed to have simultaneously a very narrow volume comfort range and much less apparent power, running out of steam and turning hard and nasty at volumes only slightly above what sounded best balanced tonally.

Adding my special grounding wire to this arrangement eliminated the gritty distortion and added back some bass extension and punch, as well as curing the imaging vagueness. However, the tonal balance was still not right, with the mid-bass and lower midrange subjectively not full enough and the mid-highs up in level, and the amps still sounded like they were 20-watt wimps.

I then started experimenting with the grounding connections of the front-end components plugged into the PS Audio Power Plants. The “balanced power” delivery of these devices notwithstanding, I soon remembered why I long ago abandoned earth grounding the front end components or the preamp. Grounding any or all of these components caused the imaging to become vague and moved the rear of the soundstage forward. This was not a subtle effect. For example, the difference in the relative positions of the DePaul University Big Band and Clark Terry on “Squeeze Me” from Clark Terry Express (Reference Recordings RR-72CD) was remarkable. Not only did earth grounding any or all of these components move the band forward to be almost as “in your face” as Terry, but the band got much louder relative to Terry as well, with the audibility of hall ambience around the band sound (which comes with distance from the mikes) strikingly reduced.

As the experiments proceeded, a pattern emerged. The effects of grounding the system in various non-optimal ways could be heard as relatively independent overlays. Grounding any or all of the front-end components to the “balanced power” PS Audio Power Plants always produced the vague imaging an forward movement of the rear of the soundstage, whether or not in combination with cheating the earth grounds on one or more of the amps. Cheating the grounds on one or more amps always resulted in a thinning of the mid-bass and lower midrange and a boosting of the lower highs, whether or not in combination with earth grounding one or more of the front-end components. Adding my special low-impedance ground wire to earth-ground an amp’s signal ground to the earth ground reference always added an additional boost to the mid-highs, together with a bit of grit to the highs.

In the end, by far the best sounding grounding arrangement proved to be plugging all the Bryston 7B-SST amps straight into the wall sockets (the recommended connection in the amp owner’s manual), with the MIT Z-1 Stabilizer staying where it was, and floating/cheating all the earth grounds of my front-end components plugged into the PS Audio Power Plants. With this grounding scheme, perceived tonal balance is now flat as a pancake, avoiding the slight brightness that was always part of the 7B-STs' best sonic signature. Bass goes down to the sub-basement with exemplary power, shudder, control, punch, and warmth. Clarity and real detail, soundstaging, and imaging are all exemplary. The grit, grain, and grunge audible with some of the grounding schemes I tried is gone. The amps seem totally unflappable; they just coast along even at extremely high volumes, with seemingly endless power reserves.

I am stunned that such simple changes in grounding can make such a profound difference in my perception of the sonic quality of a $50,000+ system which has been tweaked in every other aspect to the nth degree. While it took considerable time and effort to determine the best sounding grounding method for my system, the actual out-of-pocket cost of all the experiments and the final grounding scheme was only about $20. The sonic differences, on the other hand, were profound--the difference between unacceptable and the kind of satisfaction you can relax into. And that is almost priceless.























































Luminous Audio