Article sent to PTG on the Self-Tuning Piano
In Pursuit of the Self-Tuning Piano
Don A. Gilmore
Ever since its invention over three centuries ago, the stringed keyboard instrument—the piano and its variants—has continued to inconvenience musicians and vex technicians with the complexity of its tuning procedure. Think about it for a moment. Other musicians tune their own instruments. A child’s first lesson on guitar invariably begins with instruction on how to tune it. A trumpet player can tune his instrument any time he desires in a matter of seconds by pulling on a tube, a clarinetist by adjusting the neck of his mouthpiece. For a pianist it involves scheduling a technician to come to his home twice a year or more and tune his piano, at present a tedious process.
What is it that makes the piano so different and more difficult to tune than other instruments? One thing is simply the sheer number of strings that need to be tuned. An ordinary piano can have 250 strings or more, each of which must be tuned individually. Each string has been drawn tightly across a cast iron harp and wrapped around a tuning pin that has been driven snugly into a wood pin block. The piano is tuned by turning these pins, just like in tuning a violin. But the strings are made of high-tensile-strength steel and are strung with so much tension that the pins must be very tight in the block. This makes turning them difficult and requires much skill and experience to do so accurately.
Then, there is the matter of intonation, or the relative pitches of the different notes. In virtually all other instruments, intonation is built into the instrument or is controlled by the musician. The size and location of the holes in a clarinet decide the note that is produced, so that by closing the appropriate holes, the clarinetist is able to produce the desired tone. These holes have been permanently engineered in relation to each other and to the shape of the horn to always produce the correct tone. The frets on a guitar, likewise, are strategically spaced so that the notes played along a string will have a definite, musical relationship to each other. Even fretless instruments like the violin or the trombone leave the intonation up to the musician who controls the pitch by the placement of his fingers or hand. Unfortunately a piano is really more of a machine. It has an individual vibrating medium for every possible note. Music is played by pressing buttons (the keys), and the pianist is at the complete mercy of the instrument as far as what pitch is produced by the individually tuned strings.
Another peculiarity that exacerbates the matter of tuning pianos is the fact that modern pianos have multiple strings tuned in unison for each note. The Musical notes created by a piano are produced by a single hammer striking two or three separate strings tuned to the same pitch. For this reason, a single note can actually be out of tune with itself! That is, any of the three strings can create a discordancy.
The Obstacles to Creating a Tuning Device
For decades, inventors have tried to solve the riddle of creating a practical, easily tunable piano. It is a solution that remained elusive for a number of reasons. To be able to tune all of the strings in a piano requires the control of the tensions of over two hundred strings simultaneously. And this process must be done precisely, reliably, and economically. Patents have been issued over the years that addressed this problem in a variety of ingenious ways, employing a complex battery of individual mechanisms, levers, and motors or a system of weights to actuate each mechanism. These methods were cumbersome, complex, and costly.
Another major obstacle in creating an automatic tuning system is how to determine when a note is in or out of tune. How does a mechanism “hear” a pitch accurately? One can electronically “listen” with a microphone, but the signal is scrambled by overtones and background noise from outdoor traffic, air conditioning systems, etc. And what of the multiple strings sounding in unison? How are they to be distinguished from one another? When a piano is manually tuned, two of the three strings of each note are silenced with a rubber wedge so that only the remaining one is heard. But this is time consuming. Even with a mechanical contrivance, it would take almost as long to tune the entire piano as the conventional tuning process. And to make things worse, when a piano key is played, it is loud at first, but the sound decays quickly. This makes it difficult to determine the note’s pitch.
How Tuned is “Tuned”?
Accuracy is also a problem. Piano technicians talk about tuning accuracy in terms of “cents.” One cent is 1/100th of the pitch change of a musical half step, or from a white key to its adjacent black key, like C to C-sharp. A note is said to be “fifteen cents out of tune” if it is fifteen one-hundredths upward or downward out of tune. Two notes in unison that are detuned by as little as five cents from one another are discernibly “out of tune” by even an untrained ear. A master technician with a good ear can tune a string to within less than a cent of its correct pitch.
But what is the “correct” pitch? Without going into arcane details, the vast majority of musicians in the western world adhere to an international standard known as equal temperament and a tuning reference of A440. What this means is that the note “A” above middle-C is tuned so that its string vibrates exactly 440 times per second. All other notes are tuned relative to this note using a mathematical formula that determines what frequency they must vibrate at so that all instruments in the orchestra are playing the same notes in tune. Other standards for tuning exist, and there is much debate over which is “best.” Obviously, some standard needs to be followed.
So, with the international standard, we have a set of eighty-eight numbers (one for each note on a piano) that we can use for comparison. If we can determine accurately what frequency a string is vibrating at, we can compare this to its “correct” value and adjust it accordingly. But it is still not that simple. The sound of any musical instrument is actually a complex wave in the air that can be broken down into a mixture of many simpler waves. The dominant wave is called the fundamental, what we perceive as the musical pitch of the note. The other waves are known as harmonics and are responsible for giving the instrument its distinctive sound or timbre. In other words, if we play the note “middle-C” on a piano and then on a saxophone, we recognize them as the same note, though the instruments have noticeably different sounds.
For the piano to sound best, sometimes we need to tune strings so that their harmonics match that of other notes, regardless of whether their fundamental is exactly tuned to the equal temperament standard. This is known as aural tuning and has to do with the complex way that the human ear and brain react to the sound of a piano. Fortunately, when we do this slight “de-tuning,” the fundamental and all the other harmonics must follow. So there is still a “correct” value for the fundamental frequency to get a good aural tuning; it is just a little different than the standard one.
The Riddle Solved
For all these reasons, the path to a quick, practical, reliable, and accurate system for tuning a piano had remained elusive—until now. Over the past year and a half, I developed the solution, and it was surprisingly simple. The exclusive worldwide rights to use my system have been granted to QRS Music Technologies, Inc., and the system will be installed in a special line of Story & Clark “Prelude” grand pianos to be available later this year. The system will be installed permanently into new pianos only, does not affect their tone quality, is not visibly present, and has no moving parts.
After first pursuing the same path as other inventors investigating the possibilities of such an invention, I sought a revolutionary new approach. “It’s a shame,” I said to myself, “that one can’t just pass an electrical current through a piano string and control its pitch that way.” This reflection caused me to think further. If electrical current passes through a steel piano string, I thought, electrical resistance causes it to warm. When metal becomes warmer, it expands, and its tension decreases. When a piano string has less tension, its pitch becomes lower. Conversely, when less electrical current passes through a wire, the string cools and the pitch rises. Thus, without twisting the tuning pin, the pitch of a piano string might be changed silently, invisibly, and smoothly. Subsequent experiments have proven that this process takes place with very low voltages and with temperatures that are barely detectable to the touch.
Now, what about the “listening” problem? When a vibrating ferrous object, like a steel string, is brought near a coil (consisting of a magnetic core wrapped with copper wire), it produces a tiny electrical current in the wire that mimics the motion of the string. This is how the “pickup” on an electric guitar works. If one of these coils is provided near each string in the piano, it will sense only that string’s vibration. There can be no background noise or interference from adjacent strings. One could shout into the coil, and it would register nothing. It can only detect metallic vibrations, not actual sound waves in the air. In this way, we have an individual indication of each string’s pitch that can be monitored by an electronic circuit and evaluated.
Now, if the signal from such a coil is amplified greatly and this large electrical signal is applied to another magnetic coil located near the same string, an interesting phenomenon occurs. The second coil will actually produce an oscillating magnetic field that will physically drive the string and cause it to vibrate and sustain indefinitely, as if drawing a violin bow across it. Guitarists have used a product known as the “Ebow” since the Seventies that produces just such an effect. This solves the problem of decay and allows the circuit sufficient time to evaluate the vibration.
Nuts and Bolts
To implement this warming method for all the strings in the instrument requires that they be electrically insulated from each other at one end. If we electrically ground the harp and call it our “common” end, then insulating needs to occur at the tuning pin end. The agraffes and string rest for the self-tuning piano are made from a hard, tough, insulating material, and clear-plastic shrink tubing is applied at the end of each string, between agraffe and pin, to prevent it from touching other strings (see fig. 1).
Electrical current is applied to the strings through the tuning pins. The holes in the pinblock of a grand piano pass entirely through the pinblock exposing the tail ends of the pins from below (see fig. 2). A copper-clad printed circuit board is etched with a pattern to accept tiny, spiral-wound springs (the same as in an AA battery compartment) that solder to it in a pattern that matches that of the pins. This PC board is pressed into place and fastened to the underside of the pinblock. The copper traces lead away from these springs and terminate at the edge of the board, just under the agraffes, where ribbon cables attach and lead back to the main circuit. None of this is externally visible. Moreover, the PC board is only .06-inch thick, so it does not interfere with the removal of the piano’s action when it is serviced.
To eliminate long-term effects caused by the slipping of tuning pins, string locks are included at the string rest (much like a “nut lock” on a guitar) to effectively clamp the strings in place permanently after the initial factory tuning. This renders the tuning pins superfluous, and now only normal, seasonal environmental factors like temperature and humidity will affect the tuning.
The magnetic pickup and sustaining coils are encased in one long wooden strip that is placed directly under the strings near the dampers (see fig. 3). They are fastened to the harp from below and make no contact with the soundboard. Signals from these pickups are multiplexed onboard and run back to the main control circuit, which is hidden within the piano. The final touch is a small, gloss-black button installed in the right end of the keyslip, where its wires are run under the cheek block and back to the control circuit. This is the “tune” button.
When the piano is factory tuned, the circuit first heats the strings to a median temperature (about 95° F). This sets the “in tune” temperatures of the strings to a central value that allows them to either cool or warm when retuning takes place in the future. Then, a Story & Clark master technician expertly tunes the piano manually. Now, the fundamental frequency values for each string are stored in a memory. When the musician wants to tune his piano, he simply depresses the piano’s damper pedal and presses the “tune” button. Instantly all of the strings of the piano will begin to audibly sustain at once. During this sustaining, the control circuit polls the pickups, and each string’s frequency is determined. This value is compared with the “correct” one in the memory, and the current to the string is adjusted accordingly until all of the strings are in tune. These new current levels are now memorized and are maintained until the next tuning. The entire process takes less than twenty seconds.
Experiments have shown that to vary the pitch of an average string about fifty cents requires about one watt per string. This adds up to around 250 watts for the entire piano—about that of an ordinary light bulb. This power dissipation heats the strings to around 95° F. Since the strings are above the soundboard, all natural convection is upward and away from it. Infrared radiation effects at such low temperatures and spread over such a wide area are negligible. The system will plug into any ordinary 115 volt A.C. outlet, and it is only necessary to run the system when playing the piano. The rest of the time it is off, or it turns itself off automatically when the musician stops playing for more than a few minutes. Just before playing is resumed, the button is pressed and the strings warm up and retune.
A Threat to the Technician?
There are literally millions of pianos in homes and institutions throughout the world that are designed to last a very long time. These pianos are not going to vanish in the near future. Nor will the average piano owner leap at the chance to invest several thousand dollars on a new piano just because it never needs manual tuning. And recall that this system is only available in one model, of one brand of piano in the world. Ordinary pianos without this new innovation will account for the vast majority of pianos sold for many decades to come.
Don A. Gilmore, 38, is a project engineer and machine designer from Kansas City, Missouri. He received his B.S. in mechanical engineering from the University of Missouri-Rolla in 1986. He studied piano at the University of Missouri-Kansas City Conservatory of Music from 1972 to 1988 and has been a pianist for thirty years. He holds a U.S. patent and has three patents pending for various inventions relating to the music industry.