© Brookside Press LLC, 2012. Unauthorized use and/or duplication of this material without express and written permission from the author is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Brookside Press LLC with appropriate and specific direction to the original content.
From 1986 to 1988, I worked with electronic-music pioneer Robert Moog (rhymes with vogue), custom-building experimental keyboard instruments. In 1993, I wrote this account of our work together, and an abbreviated version was published around that time in Piano & Keyboard magazine, no longer in business. This is the first publication of the unabridged version. A Postscript with updated information about Moog and his family follows the article. — L.F.
When Bob Moog called me in January 1986 to ask if I would work with him on a small project, the last thing I needed was another job. I was running a piano-service business, finishing up work on The Piano Book, writing a regular monthly magazine column, and doing about ten hours a week of charitable volunteer work. I couldn’t see fitting another activity into my schedule, so I said no. Then, after hanging up the phone, I thought, “Larry, you fool — how often do you get the opportunity to work with someone of this caliber?” So a few days later, I called back and said maybe.
Several weeks later found Moog and me standing in his freezing garage, inspecting a Yamaha CP-80 electric grand. I had serviced these hybrid electric/acoustic pianos before, but only occasionally. Their distinguishing feature is that though they have no soundboard and are electrically amplified, they do have tuning pins and strings, and a real grand-piano keyboard and action.
We had removed some of the outer case parts and were peering at the action. “Do you think you can remove that?” Moog asked, gesturing at the action, in what sounded like a cross between a genuine query and a test question. After all, Moog had found me because, like him, I was a columnist for Keyboard magazine. “He can write,” I could imagine him thinking about me, “but does he know anything about pianos?” Of course, removing a grand action stack is something any self-respecting piano technician can do blindfolded; I unscrewed eight or ten screws and out it came. Moog appeared impressed; though I was surprised by his response at the time, I don’t know why I’d assumed that this famous electronics engineer would necessarily know much about piano actions. Anyway, having clearly passed the test, and since Moog said the project would take only about three months’ worth of Saturdays to complete, I agreed to work with him.
Had I known then how complicated the job would be, and that it would last not three months but two-and-a-half years’ worth of Saturdays (and sometimes Sundays, plus three months of full-time work), I never would have signed on. It’s a good thing I didn’t know, because it turned out to be one of the most fascinating and enjoyable jobs I’ve ever had.
I’m sure Moog must have explained to me on that first day what the project was all about, but looking back, I can see that it was months before I understood it or saw the whole picture. Apparently, for years Moog had been in consultation with musicians and composers, most notably John Eaton, then of Indiana University, on ways to expand the limits of keyboard instruments. Over the years, tone-producing technologies had advanced by leaps and bounds, but keyboards had changed little from their traditional design. On an acoustic piano, for example, a musician has control over only three parameters: which note is played; how long it sustains (and even then, only until the sound dies away); and the velocity with which the key is depressed, which governs the loudness. On some electronic keyboards, only one additional parameter has been added: polyphonic aftertouch, the pressure of the key at the bottom of its stroke, the sound of which depends on the particular keyboard or is programmable by the user (not to be confused with a piano’s aftertouch, which is something entirely different).
It was Moog and Eaton’s idea to expand the number of different operations a player could perform at the keyboard, and to make each of those operations programmable by the user. For a long time this idea remained only a dream because of several technological obstacles: computers were too slow and expensive, sensor technology wasn’t advanced enough, and interfacing with tone-producing elements was difficult due to lack of a common computer language for musical instruments. By the mid-1980s, however, all of this had changed with the advent of cheap personal computers, further miniaturization of sensors, and the development of the Musical Instrument Digital Interface (MIDI).
Moog planned to add several features to the traditional keyboard. First was a touch-sensitive keytop (playing surface) that would sense, for each key, the position of the player’s finger from left to right and fore and aft, as well as the total surface area of the finger on that key. Another new feature would be a sensor that measured a key’s vertical position in its stroke. By computing that position over time, the key’s velocity could also be determined. Finally, a force sensor would be included at the bottom of the keystroke to provide the usual aftertouch-pressure information. Each of these operations was to be assignable to any aspect of tone contained in the MIDI specifications, such as loudness, pitch, vibrato, tone color, reverb, and many others.
For example, the vertical position of the key could be used to control the loudness of that note while the amount of aftertouch pressure controlled the degree of vibrato — and the surface area of the finger on the key controlled some aspect of the tone’s harmonic characteristics. From the virtually unlimited number of possible combinations (subject, of course, to the capabilities of the tone-producing devices to which the keyboard is attached), the player would program his or her choices into a computer connected to the keyboard, and these choices, which might change over the course of a piece of music, would become part of the composition itself. The name to be given to this odd instrument was the Multiple-Touch-Sensitive (MTS) Keyboard.
Actual construction of the MTS keyboards had briefly begun at Moog’s workshop, next to his home in North Carolina, several years prior to our meeting, but everything had to be hurriedly packed up and moved when Moog accepted an invitation from Ray Kurzweil to become Vice President of New Product Research at the newly formed Kurzweil Music Systems, near Boston. While the Moogs spent a year adapting to their new home and life in the Boston suburb of Natick, most of their keyboard gear remained packed away in the garage, where we found it on that cold February day. In addition to the Yamaha, which was to be turned into an MTS keyboard for New York musician Gregory Kramer, there were four four-octave organ keyboards (three for John Eaton, and one for Moog to experiment with), and one six-octave keyboard for Steve Porcaro of the band Toto, an order that was later canceled when Moog realized he would not be able to fulfill it within a reasonable length of time. Moog decided to begin with the Yamaha, so we separated the keyboard part from the strung back (easily done on this instrument) and set the keyboard up on its legs in the shop.
Moog’s “shop” in his Natick home was a far cry from the ample industrial building he had erected in North Carolina. It was actually a large furnace room with a concrete floor, perhaps 10 by 20 feet, with the furnace and water heater at the far end. The rest of the room was quite filled up: along one long wall were two large workbenches and a couple of filing cabinets; along the opposite wall, rows of steel shelves extend to the ceiling, stuffed with every manner of industrial and electronic hardware; with a third workbench and drill press near the door. This left a long, narrow space for the keyboards and us. The relative lack of space was a source of some frustration for Moog, but, fortunately, the two-car garage accommodated some overflow, including a radial arm saw, a belt sander, and some additional work space. Next to the workshop room was a more spacious office, with desks, computers, and a Yamaha upright piano. Both of these rooms and the garage were on the entry level of the house; the family’s living space was upstairs.
The major activity in building the MTS keyboards was the fabrication, wiring, and installation of the keytop sensors. The keytops for the naturals were cut out, oversized, in the shapes of their respective keys from thin sheets of epoxy-glass circuit-board material, each containing several octaves’ worth of keytops. On one side of each keytop had been laid a conductive pattern, leading to terminals at the four corners. The keytops for the sharps (piano technicians call all black keys “sharps”) were small rectangles of such material, also cut from sheets. On each keytop, on the side opposite the conductive pattern, after I carefully masked off areas not to be painted, Moog screened a thin film of black “resistive” paint (i.e., paint that conducts electricity but with some resistance). The paint was cured under heat lamps, then sprayed with a thin coat of urethane. When everything was dry, I soldered a multi-wire ribbon cable to the terminals at the back end of each keytop, and then a connector plug, for connection to scanning circuitry, to the other end of each cable.
The way the keytop sensors worked was described to me this way: The resistive paint surface and the player’s finger form two plates of a capacitor, the finger being considered grounded at high frequencies by virtue of its connection to the rest of the body. The urethane coating over the resistive surface is the insulating dielectric of the capacitor. A high-frequency alternating voltage is applied to the four corner terminals and the painted surface via the ribbon cable and the conductive pattern on the back of the keytop, and the resulting current at each terminal is measured. The proportion of the total film current measured at each corner terminal indicates the position of the finger relative to the corners. Each key is electronically scanned 200 times per second to give a continuous reading of these values.
After wiring up the keytop sensors, the next step was to glue them to the keys. On the Yamaha we decided to glue them directly to the plastic keytops already on the keys. Removing the plastic seemed unnecessary and difficult, and these being one-piece tops and fronts, would have left us without key fronts as well. First we had to trim a small piece off the back ends of the Yamaha keytops to accommodate the ribbon-cable connections. Each sensor was then glued to its appropriate key with five-minute epoxy and clamped until dry.
Following the gluing came the most tedious and time-consuming job of the entire project: trimming and filing the oversize sensors to exactly match the shapes of the keytops to which they were glued, as well as to eliminate any sharp edges and create a uniform appearance from note to note. Despite the tedium and epoxy dust, I found this job strangely satisfying, probably because some semblance of art was involved. After testing for good connections, the black-painted keytop sensors on the natural keys were painted with white epoxy to once again resemble a piano. The bottom edge of the fallboard was trimmed to accommodate the now slightly higher keytops.
The 88 ribbon cables now trailing from the keys had to be connected to the scanning circuitry, which consisted of 11 circuit boards. The question was where to put all this stuff without fouling up the movement of the keys and action — one reason Moog had hired a piano technician to assist him. Fortunately, almost as if anticipating our need, Yamaha had thoughtfully provided a rather large, empty space beneath the keyboard of the CP-80. Moog cut out a portion of the instrument bottom, hung a couple of hinged trap doors, and installed the circuit boards inside. A shallow slot was cut down the side of each key, and each ribbon cable was dressed down the slot into the cavity below. The cables were carefully routed so that the collective mass of wires would not push up on the keys above, thus limiting their movement. The 11 scanning circuit boards were all wired together and connected to still other circuit boards that made sense of their data.
As mentioned, Moog’s other innovation was to add a sensor for the vertical position of the key in its stroke. This sensor, another variable-capacitance device, consists of an aluminum vane attached to the bottom of the key and a pattern on a circuit board mounted on the keybed below. In this case, air is the insulating dielectric and the capacitor’s output depends on the distance between the vane and the circuit board. The vanes — small rectangles of thin aluminum — were stuck to pieces of foam rubber and attached to the bottoms of the keys, carefully positioned over the circuit-board patterns. The boards were spring loaded so that their distance from the vanes could be finely adjusted. As with the keytop sensors, these 11 circuit boards were also wired together and connected to other circuit boards in the cavity below.
(continued on next page)
To comment on any article, go to Larry's Blog.
Brookside Press LLC
P.O. Box 4916
Palm Springs, CA 92263 USA
Copyright 2014 Brookside Press LLC.
All rights reserved.