Mix Interview: Arthur KelmA WELL-GROUNDED APPROACH TO AC POWER PROBLEMS 6/01/2009 8:00 AM Eastern
Electrical power, like the air, is something most of us take for granted. Live or studio, AC power and grounding remain one of the great enduring mysteries in audio systems. Hum, ground loops and intermittent buzzes seem to be prevalent in studios everywhere, and there are few guitarists who haven't encountered the guitar-to-mic “kiss” stemming from improperly wired outlets in clubs or rehearsal spaces.
Solid information about power and grounding in audio is sketchy at best and shrouded in folklore, myths and the ever popular “I read this somewhere on the Net, so it's gotta work.” However, one person who's taken an empirical approach to the subject is Arthur Kelm, studio designer and CEO of Ground One AV Inc., a consulting firm that addresses the specific power and grounding requirements of pro recording facilities and high-end home theater installs.
Kelm's audio roots run deep. After getting his EE degree, he worked on satellite communications systems and then left to create custom wireless rigs for (and toured with) Debbie Reynolds. Coming off the road, he was chief tech for Chateau Recorders in North Hollywood and designed Broad Recording Studios for Hawaiian artist Al Harrington with the help from his mentor, legendary studio architect Jack Edwards. Returning to L.A., he spent days as a tech at Canyon Recorders and nights at Record One Studios. This was followed by a couple years doing engineering and tech work for Laura Nyro (for whom he designed a studio), Toto and Ocean Way Studios. In 1987, Greg Ladanyi and George Massenberg asked Kelm to be general manager/chief engineer of The Complex, and he remained there for four years.
Kelm then provided freelance tech/design services and consulting for companies such as Walt Disney Imagineering, and was director of engineering for The Record Plant Studios. After a year as chief engineer at Skywalker Sound, Kelm focused his energies full time on studio design and creating solutions for AC power problems.
Today, Kelm's company offers consulting services and manufactures AC distribution, voltage conditioning, power isolation and filtering products. He also still takes on the occasional studio design project and graciously set aside time to discuss some design/construction/installation and grounding issues.
It's first determining the artist's needs and then making that artist's favorite speakers work in that room design. I learned studio design from hands-on experience creating a lot of rooms, seeing what worked and what didn't work. Being a technician helped me see all the elements that go into a good room design — it's not just architecture.
You can design a room entirely from drawings and computer models, but that's no guarantee the room will sound good. And one room might sound good with certain speakers but not with other monitors. And 90 percent of the time, that's the issue — the wrong speakers in the wrong room. A good room should sound good — period — but the wrong speakers can make a great room sound bad and the speaker/amplifier combination is critical. There's no such thing as a perfect monitor.
It's only complicated because of grounding. If it's not done correctly, there are a lot of red herrings. People will begin troubleshooting a hum or buzz and start looking in the wrong place to fix a problem. Then when things start interacting with each other, it begins a downward spiral when you may “fix” one problem only to have it move somewhere else, which screws people up. I take an analytical approach to solving power issues. People try all sorts of odd remedies, when sometimes the problem can be traced to poor installation by the electrician.
Some techs feel power and grounding is a lot of black magic that doesn't make sense, but to me it's not that complicated, especially with only three, four or six wires. I was recently working on a few home installations where they had problems. I found sump pumps, washer/dryers and air-conditioning units on the same panel board as their sensitive electronics and computer gear!
In studios, I've seen just about every type of home-brew solution, like people driving ground rods under consoles. Then I hear comments like, “The system works, but some days it buzzes and other days it doesn't.” It comes down to trying to use the audio chain to compensate for something that wasn't done properly in the electrical chain. For years, I didn't look at the electrical side when trying to solve hum and buzz problems; I only looked at my world. I'd use a whole box of Band-Aids, gum and baling wire — lifting a ground here or inserting fiber washers under the rack screws to isolate a piece of gear somewhere else — to get problems to stop.
Eventually, I decided to reverse-engineer the process. Once I read more and talked to people in telecommunications about power and grounding, I got a good understanding of it. I bought my first Fluke 43 [power-quality analyzer] for $4,500, worked with it and a light went on inside me. Testing and test equipment actually works with power systems, and if you do it right and follow the NEC [National Electric Code] book, it ends up quiet and clean.
What's your approach to studio grounding?
I use a star grounding technique, but to me an isolation transformer is mandatory. The key to what I do is re-bonding the neutral and ground on the secondary of the isolation transformer. That feeds your panel board, and the panel board feeds all the outlets in your studio. When you re-bond neutral and ground on the secondary of the transformer, you are actually, in terms of the electrical code, creating a new service, and 90 percent of your noise — click and pops and other disturbances — comes from voltages between the neutral and ground. So once you've re-bonded neutral and ground, you're off to a new start.
I do soil-resistivity testing to determine what kind of ground reference I have. I shoot for under 5 ohms path-to-ground, and in some cases it's necessary to drive a secondary ground rod. The code is 25 ohms for a service entrance, where the electrical service enters the building and where they bond neutral and ground. The 25-ohm figure is for life safety, but you want to be under 5 ohms for a low-noise ground.
If I don't have a great ground at the service entrance, I'll do a supplementary secondary ground after the isolation transformer. The ground from the service entrance comes to the transformer, goes to a bus bar, the neutral jumpers over to the bus bar and then I'll go from that out to a brand-new ground rod or rods — to get under 5 ohms.
That becomes our new point source for ground. It goes from that point up to the sub-panel, and the sub-panel has an IG [isolated ground] bus in it and an isolated digital bus in it. Then all the neutrals and green wires from the receptacles come back to that point — you never daisy-chain anything. It's a star system from that aspect. Each receptacle or quad box has a dedicated run back to the panel and to its own breaker. That approach has worked flawlessly for me for the past 10 years.
In terms of the ground rod itself, is a standard Home Depot-type, 5/8-inch copper-clad steel rod okay?
Maybe, but you have to make measurements. Ideally, I like to use an ionic electrode [chemical] ground rod, like the Lyncole XIT. It's a 2-inch-diameter tube with weep holes drilled into it and it's filled with a rock-salt material. You core a 6-inch hole, place the rod in the center and backfill it with Bentonite clay/soil conditioner, which you mix into a slurry with the thickness of pancake batter. The Bentonite creates a high-conductivity/low-corrosive mixture.
It's all got to be calculated. I go out to the site, pound stakes into the ground and make Wenner soil-resistivity measurements in 5-foot increments, from five through 20 feet. From that I make a soils profile based on the continuity of the soil and can calculate the required length of the ground rod to have a 5-ohm path-to-ground. This can vary from a single 10-foot chemrod to two or three 15-foot chemrods. When I did O'Henry Studios [Burbank, Calif.], I needed two 20-foot chemrods to achieve five ohms. The facility was in the [San Fernando] valley in very sandy soil, so we needed to go that extra mile. Also, when ground rods are placed too close to each other, they act as one — rather than separate — ground rods. If you're installing 10-foot rods and need two, you need to place them 20 feet apart so they don't interact with one another.
If you have the money, go for a chemrod — they're expensive but last 28 years. The alternative is driven-steel rods, but they lose their effectiveness after about 10 years as they rust away. For all intents and purposes, they're just pieces of copper-plated rebar, and that copper coating wears away as the rod is pounded into the ground. Once the copper's gone, the bar starts rusting.
Ground maintenance is also important. You need to check your ground every few years with a clamp-on meter. Only about 10 percent of the ground systems I've checked even met the code — which is 25 ohms — and most measure somewhere between 35 to 600 ohms.
Oddly enough, if electrical inspectors see a water pipe ground and a driven ground rod, they won't even bother to measure it, saying, “That's good enough.” It can change with time and soil conditions. Moisture can help, but the quality of the ground is based on the soil's mineral content.
Once you have a good grounding system established, it's amazing how much better things are. It really works and it's not rocket science.
So if humans can fly to the moon, why can't I get the buzzing out of my Strat?
Blame it on the Earth's magnetism. [Laughs] Actually, it's not the Earth's magnetism, but unless you get way out in the country — like out at Neil Young's ranch — you're constantly being inundated with magnetic fields. And single-coil pickups are the most sensitive device we have in the audio business. If there's a hum or magnetic field out there, they'll find it. Concrete/steel rod-reinforced buildings are huge Faraday cages — they actually radiate, bringing hum into a system. They're giant 60Hz antennas.
In some outboard devices like AMS units and Harmonizers, their power supplies throw off a huge [EMF] field, which you can prove by putting your guitar close to them. I have a single-coil pickup and Danelectro HoneyTone battery-powered amp I can turn on and move around inside buildings and track down hums. When I design a room, I try to place all electrical within steel conduit and twist the hot and neutral wires together to cut down on EMF fields.
Lighting dimmers can be culprits, and APC UPS [Uninterruptible Power Supply] units can back-feed noise into the electrical system.
What advice do you have for the budget home studio?
Always have a single source for your audio power. In most bedroom/home studio conversions, one of the biggest issues is having two separate circuits that are daisy-chained back to the panel through some other rooms. Here, you're destined to have a hum or buzz if you wind up having two paths back to the panel board of varying length, with different things plugged in along the way. Most small rooms will have a single 20-amp circuit, so go from that to power strips and avoid plugging audio gear into different outlets around the room.
If you need to go beyond a 20-amp capacity, you should call an electrician in to pull in a separate circuit — or two dedicated circuits in that room. Going slightly more upscale, have an electrician run a 240-volt circuit to the room and go from that to my rackmount isolation transformer, which splits it to six 20-amp breakers. Here, the isolation transformer separates you from the house, creating your own small service.
You can designate certain circuits in a room as “dirty” and use those for lighting and non-audio gear.
Exactly. If you have three 20-amp breakers feeding a room and put all your audio gear on one breaker and everything else on another, you're already miles ahead of where you were.
Any surprises you've learned along the way?
Just because you've hired a good chief electrician doesn't mean the workers on the job pulling wire and putting in plugs are as good. I had this situation doing Don Henley's studio installing an old API console, which was quiet. After putting in three pieces of outboard gear, the noise floor started moving up. I turned off the power, went to the main panel board, pulled all the green wires off the IG ground bus and started testing them back to the utility and found four shorts. On the neutral bus I found the same thing — five more shorts, shorted to utility. I called the electricians and told them there were shorts on the IG outlets. They begrudgingly came out and fixed the shorts, and everything was dead-quiet. Oddly enough, these faulted outlets were in other parts of the studio I wasn't using. Never assume anything with electricians.
George Petersen is the executive editor of Mix and runs a small record label at www.jenpet.com.