Why does my amplifier hum? Because it doesn’t know the words.
Old joke, but to the point. Electronic gear sometimes makes noise all by itself—not noise we hear from the speakers, but noise we hear in the room. Live noise, if you will. Even at low levels, it’s annoying; that continuous 60Hz hum is never in tune with anything you’re playing. And if you have a mic open in a roomful of hum, it’s all over (in both senses of the phrase).
So this episode of the ‘Shut Up’ saga concentrates on hums and how to get rid of them (sorry, Pooh). I’ll spend most of my time on mechanical noise, but I’ll also touch on the subject of inductive coupling and report some tests I’ve done on a few popular microphones, measuring their susceptibility to induced hum.
As a rule, the worst hummers are gadgets that use a lot of juice from the wall; the bigger the transformer in the power supply, the worse it’s likely to hum. Most of the culprits are power amps, although there are exceptions (NB has a synth power supply that has been singularly hard to silence). Here’s a typical example.
The Case of the Humming Dynaco
In the course of writing equipment reviews, I’ve occasionally described my monitoring set-up: a pair of Rogers LS3/5a nearfield minimonitors driven by an aged Dynaco Stereo 70 power amp. The Dynaco, a moderately powered tube unit, is a synergistic match with the speakers (designed by the BBC for their remote monitoring trucks). Within the limits of the speakers’ passband (essentially 90Hz and up), the combination provides a beautifully detailed, neutral-toned sound that tells me much of what I need to know about my recording. (No, not everything, and I’ve been bagged a few times.)
Unfortunately, the hefty power transformer on the Dynaco hums. I use it on remote recording gigs, with the amp parked strategically under the table (it makes a nice foot warmer, about as effective as a small cat). Over the years, I’ve tried various ways of isolating the amp from the floor.
In most rooms, placing the power amp on a pair of old tennis shoes, while unsightly, is reasonably effective. In a few places with poorly sprung floors, I’ve rolled up bath towels and placed the amp on them, taking care not to block the ventilation holes.
But my new control room posed a bigger challenge. It’s pretty cramped, and there are few locations suitable for the Dynaco. The most effective, unfortunately, is the worst from the point of view of hum: sitting atop a stack of other components on the right side of my desk.
Why is this location so awful? Each component has a lid that resonates beautifully at 60Hz, and the table top amplifies the resulting cacophony. Even after applying some of the remedies outlined in Part 1 of this series, the room still hummed like Allen Ginsberg chanting “OM” (so long, Allen, and thanks for the fine poetry). What to do?
I have a repertoire of desperate measures. I tried felt pads (old mouse pads, typewriter pads—remember typewriters?) and bean bags (made from dried beans stuffed into old socks). No soap. I contemplated setting the amplifier in a small sandbox made from a developing tray. It’s worked for me on a temporary basis in the past, but I envisioned sand getting into every corner of the room, including tape transports. No.
The grommet comet
Instead, I decided to attack the hum at an earlier stage: the point where it was coupled from the power transformer to the amplifier chassis. This particular hummer is mounted horizontally, with bolts running through the body of the transformer and thence through holes in the chassis. Nice and solid—and a great conduit for vibration.
My remedy is shown in Figure 1. After making sure the amp had been unplugged from the wall for at least an hour, I unthreaded the nuts and lock-washers from the bolts, removed the transformer from the chassis, and threaded soft rubber grommets onto the shafts of the bolts. (I bought the grommets at my local electronic surplus outlet. Many hardware stores sell them, too.)
Next, I lowered the transformer gently into place and threaded more grommets onto the ends of the bolts as they poked through the holes in the chassis. I threaded the nuts back onto the bolts (without the lock-washers, which were no longer needed) and tightened them until the grommets just began to compress.
I put the amp on a large hollow table and plugged it in. Dead silence; not a hint of hum in the room. I moved it back to the stack of gear; still no noise. I pronounced the amp cured. Total cost, 80¢ plus tax.
I was lucky: the holes in the chassis were bigger than the bolts on the transformer by a sufficient margin that I was able to keep the bolts from touching the chassis, which might have ruined the isolation. If the holes had been smaller I would have enlarged them with a drill or a sheet metal punch; under those circumstances, I’d be tempted to open out the holes to the point that I could insert the grommet directly into them.
This works for horizontally mounted transformers. Vertically mounted units use brackets and non-captive bolts to fasten the transformer to the chassis; they can be isolated by floating the transformer between a pair of grommets (see Figure 2).
Although decoupling the transformer is the best solution, you may find that isolating the component does the job adequately. I outlined a few methods above (tennis shoes, felt pads, etc.). There are more possibilities.
AudioQuest sells thick, rounded equipment feet made from Sorbothane, a boingy substance with a texture disconcertingly close to human skin. (It was developed originally for burn treatment.) The feet come in two sizes: CD Feet ($39.95/set) for lighter weight equipment, and Big Feet ($59.95/set) for heavier gear like power amps.
Pricey, but in my experience they do a good job of decoupling mechanical hum and other vibrations; the improved isolation can also have surprisingly beneficial effects on the sound of equipment. (Everyone knows vacuum tubes are microphonic, but so is a surprising amount of solid-state gear.) AudioQuest products are available from Audio Advisor (800/942-0220). Other similar products are available at many stereo dealers, but I can’t tell you which ones work best because I haven’t tried them yet.
If you don’t care about looks, you can try setting a problem amp on a pair of small sandbags, being sure the bags are tightly sealed. Another possibility, which I haven’t tried, involves the small rubber balls filled with something squeezable (sand? pebbles? buckshot?) that are sold as exercise devices at drugstore checkout counters. They seem to be very dead, and might serve as good isolators. Tough, too. If you try them, let us know the results.
On the rack
Rack-mounted gear presents special difficulties in isolation; often the equipment is attached to the rack only by the front panel ears, which is an awkward place to use grommets as decouplers. In cases like this, I’d try to isolate the transformer first; if that fails, I suggest removing the offending piece from the rack, or mounting a shelf in the rack and setting the gear on the shelf using Big Feet or the equivalent.
Off the wall
“Wall-warts” (small, plug-in transformers with umbilical cables running to pieces of equipment) have come to plague the modern studio. Manufacturers like them because they can be bought from outside makers pre-certified by Underwriters’ Laboratories (getting UL to certify a piece of new equipment is an expensive process). Because they’re external, they also remove a source of inductively-coupled hum from the chassis.
Unfortunately, if you plug a wall-wart into a power strip mounted in your rack, its mechanical noise can couple to the rack, and you’ve got another hummer on your hands. (They can even excite vibrations when plugged into wall sockets if the wall is flimsy enough.)
The best solution is to soft-mount the wall-wart. You can get (or build) very short extension cords that let you plug in the wall-wart without a hard mechanical connection to the power strip. Laying the transformer pack on something soft or springy usually provides adequate isolation.
Certain pieces of gear (not all of them power amplifiers) have problems with electromagnetic radiation. I have one wall-wart (used to power a set of wireless headphones) that radiates a vicious hum field; it’s capable of inducing audible noise in the heads of a cassette deck from four feet away.
Such units need to be located far away from sensitive devices (transformer-coupled mic preamps and analog tape decks); there is no effective home-shielding mechanism of which I am aware. Extension cords can help remove wall-warts from the line of fire; larger items like power amplifiers may need to be relocated away from sensitive gear.
Among the devices most sensitive to inductive hum coupling are microphones. In the process of reviewing mics for the magazine, I devised a simple test that yielded some thought-provoking results; for this article I checked out every microphone I had sitting around the house. My field source was the horrible wall-wart; my method was to place each microphone with the center of its body located 2" from the transformer, rotating the unit on its axis to find the worst-humming orientation.
I ran the signals through my mic preamp (matching sensitivities with the gain controls) and thence to my DAT machine, using its headroom indicator as a meter. The results are on an arbitrary scale, but relative susceptibilities can be compared. In each case I measured the microphone with several parts located closest to the transformer; Table 1 lists the worst-case results (and notes which part of the mic was most vulnerable).
Do you see a pattern? With occasional exceptions, transformerless condensers were the least vulnerable (two of these had no measurable hum at all), followed by transformer-coupled condensers, followed in turn by dynamic mics. This makes sense; transformers inside microphones are small and poorly shielded, so they are easy prey for nearby hum fields. And the capsules of dynamic mics can pick up hum, even if the body is adequately shielded.
The principal use of these numbers is in choosing a mic for recording combo guitar amps. These can throw a respectable hum field, and I’ve occasionally had problems recording with dynamic mics. Short of switching to a piggyback amp (and moving the head away from the speaker and microphone), the most obvious solution is to pick a mic with good hum rejection. I’ve found that the SM-81 is a good choice for recording problem amps, and in practice the RE-200 or even the RE-15 is usually fine except in the worst cases (I remember an old Supro that was a nightmare).
Wrapping up and Shutting Up
We covered a lot of territory in this series: resonances in the room, garbage in the electric wiring, and now the vagaries of hum. There’s plenty more to talk about: soundproofing the room to keep out the noises of passing cars, planes and birds (I just spent ten minutes editing a new final chord onto a piece where the ringout faded into a nice background of tweet-tweet). And the subject of getting air into a studio while keeping HVAC and other noises out is worth an article in itself.
But those subjects are for another time and another series. Until then, happy silence!
Paul J. Stamler is a producer/engineer and session player in the St. Louis area.
Paul J. Stamler is a producer/engineer and session player in the St. Louis area.