Pianos are complicated pieces of machinery. Contemporary pianos have literally thousands of moving parts; it’s mind-blowing when you think about it. And it’s amazing how well the design works: the resulting instrument is fluid, expressive, dependable, extraordinarily loud for a string instrument, musically versatile, and a pleasure to play and listen to. This is not surprising when you consider that 300 years of diligent thought and effort have gone into the perfecting of it. Yet I have often found myself thinking it ought to be possible to design and build a much simpler piano-like instrument. Not one that could compete with the modern standard model in the above-mentioned qualities, but one that would at least be easier to build in a home workshop. Also, one that might have a different sort of sound, quieter and less commanding; perhaps more like a lute or guitar.

So I set myself this task: to design and build the simplest keyboard zither I could come up with – one which might have just a couple of moving parts per key. The minimum requirement would be that the note sounds when the key is depressed and stops when the key is released. The range could be much smaller than the piano’s seven octaves, and it would have one string per note instead of the piano’s three. I gave this project the name Minimal Clavier.

This could be seen as a foolish idea. Why should I knock myself out trying to make a functional home-built keyboard string instrument when the world is already full of such instruments built to a standard far higher than anything I could dream of? One of the advantages of making unconventional one-of-a-kind instruments, as I usually do, is that once you’ve completed a one-of-a-kind you’ve just made the best one in the world. Whatever I might come up with keyboard-wise would most assuredly not be the best in the world. But I did like the idea of a small, portable keyboard, light in both weight and sound, filling a different niche from the standard instrument.

We should note here that there have historically been many smaller and simpler keyboard string instruments, and many of these are still made in replica. This includes a variety of early pianos, harpsichords and clavichords, many of which were indeed small and light and simple in action. There are even some 20th-century oddities such as the Dolceola – you can look that one up if you’re not familiar with it. And there are, or were until not so long ago, more or less conventional but somewhat smaller pianos, sometimes referred to as “railroad pianos.” So, to repeat, it does look a bit foolish to think of making what would likely end up as a half-assed version of something which already exists in fine form. Still I persisted.

My original intention was to make an actual piano: an instrument sounded by a keyboard action which throws a hammer against the string and allows it to bounce off. But I also remained open to the idea of a harpsichord-type action, which would involve plucking rather than hammering. Eventually I settled on the plucking action, both because I preferred the sounds I was getting in preliminary experiments, and because I was able to arrive at what seemed like a simpler mechanism that way. But in case anyone’s interested in the hammering approaches I considered, here are a few notes about that. First, be it noted that if you’re looking for a really simple hammer-throwing action, a good place to start is with old-fashioned toy pianos. They’re not string instruments, but the design of the action could be applied to strings. They represent the hammer-throwing idea reduced to its simplest form, with just two independently moving parts per key. They don’t include dampers, though. For my piano action, I had a different thought. The idea was to place the hammerhead at the end of a springy metal arm. When you press the key, the farthest extent of the key-lever movement would stop short of string contact, but the flexing of the metal arm would allow the hammerhead to continue to fly forward, strike the string and bounce off. The difficulty with such actions is, the springiness tends to cause the hammer to then swing back and forth, hitting the string repeatedly. This is what we see in the early 20th century instrument called Marxophone, where the repeated bouncing of hammer on string is regarded as a feature, not a bug. But in a would-be piano, repeat strikes are unwanted. I tried various tricks to stop this from happening, but never got it fully in control. For the damper I had another small armature extending off the far end of the key lever going around behind the string, with a pad affixed opposite. With the key in rest position the pad would rest against the string. Pressing the key would cause it to lift, leaving the string free to sound when the hammer hits. 

As I’ve said, in the end I decided not to go with a piano-like hammer action, but instead with a harpsichord-like plucking action. It’s not hard to envision a very simple keying mechanism in which the far end of the key has both a plectrum and a damper, positioned so that the same motion causes the damper to lift and the plectrum to pluck. But there are some challenges here. The most vexing one is, how does the plectrum get back to its starting position in front of the string when the key is released, without plucking again on the way back? You can go online to find the solutions that have evolved through generations of harpsichord makers. They typically involve having a couple of independently moving parts which allow the plectrum to fall back out of the way as the key returns. Still, even in well made harpsichords, you can usually hear a small noise when the key is released. People seem to find this small end-of-note sound easy to overlook. (I actually kind of enjoy the key-release sound you sometimes hear as the last sound in a live or recorded piece of harpsichord music.)  In any case, these well tested systems are highly recommended for anyone that wants to make a really well-functioning harpsichord (and who has in generous measure the skills and patience required, not to mention tools and space).

Side view of a typical key in the Minimal Clavier

For my Minimal Clavier, I went for something simpler still, since that was the point of this project. You can see what I came up with in the diagrams shown here. (Feel free to skip this paragraph if you’re not interested in the details.) The keys, with the plectrum and damper at the end, are mounted on bands of spring steel (1/2″ x .062″) so that you can press the wooden part of the key downward and it will spring back when released. The damper functions as you’d expect, being a sponge pad on a little wire armature at the end of the key that goes around and behind the key so that the damper normally presses against the string, but moves down and away from the string when the key is pressed. 

Detail of the plectrum and damper

The plectrum is a piece of very thin spring steel (.010″ x 1/8″), shaped and bent at the end as shown in the lower drawing so that the short bent section does the actual plucking.  The original hope was that the long arm of the plectrum would flex backward when the key returns, allowing the plectrum to slip past the string with minimal re-plucking sound immediately prior to the return of the damper.  But in practice this system didn’t work so well. I found that if the plectrum material was thin enough to easily flex back for the return, the very thin material would tend to engage in a sort of forward-flexing during the initial movement which would jam rather than plucking – in short, if the plectrum piece was good at flexing out of the way on the return, it was bad at plucking in the first place. I resolved this by fixing another straight metal piece – same 1/8″ width as the plectrum but thicker and more rigid — in front of the long part of the plectrum piece. The rigid added piece prevents the long part of the plectrum from flexing forward during the pluck, but still allows it to flex backward during the return.

For the most part, this approach works: when the key is pressed the damper moves away and the note usually plucks clearly. On the return there is a bit of noise as the plectrum hits the string and flexes aside just before the damper returns to stop the string, but this noise is not much worse than it is on other harpsichords. But let’s be honest about this: this is not the sort of 100% dependable, easy and smooth action you’d want and expect in a well made keyboard. I’ll discuss the shortcomings more fully in a moment, but first a bit about the zither – that is, the string-covered box that is the sounding body.

The box itself is a rectangle of about 24″ x 21″ x 3″. The construction is straightforward for a soundboard and sound box.  It’s pretty, because for the back and sides and some other parts I used the wood of a black acacia tree that had to be taken down in my yard, and acacia is an attractive wood. There are 30 single strings for a chromatic range of two and a half octaves. Both bridges are located on the soundboard, which helps to give it a warm sort of sound that I like. (In theory this would make the tone favor the odd numbered harmonics.) The board itself is just 1/8″ plywood, with struts to help it take the stress of 30 strings – but the string tensions are kept fairly low in any case. The strings are steel, wound for the lower strings and plain in the upper. The key levers are made from laminated bamboo fiber flooring that I happened to have around (very hard and stable), topped with ipe for the nominally black keys, maple for the white. The look of the elegantish acacia box with the steam-punkish keying mechanism on top is a little odd. The tone of the zither alone when plucked by hand is pleasant enough, and reasonably loud, but the tone when played with the keys more impressive – surprisingly loud and full. That’s partly just because the plectra are activating the strings quite strongly. It’s also because of the direction of the pluck, exciting the strings in the direction perpendicular to the soundboard. This plucking direction (which happens naturally in pianos and harpsichords) makes for efficient transmission. In this particular design it should also help the odd-harmonics effect to come through.

And now to discuss the problems with this instrument, hopefully in a way that will be instructive for any who might be curious themselves about making a simple harpsichord-like instrument. The problems all have to do with the plucking action. Even after tweaking every tweakable adjustment in hopes of making the individual keys work well, the plucking action on this instrument still tends to be inconsistent, unreliable, and overly stiff, making the instrument difficult to play. The problem is that tiny differences in the way the plectrum contacts the string make a big difference in playability. The difference between a smooth action and glitchy one may lie in a fraction of a millimeter’s variation in how close to its tip the plectrum crosses the string. A tiny bit too far one way and the plectrum is in danger of hanging up on the string rather than plucking. A little less and the string may sound, but still too aggressively, making a harsh tone with a very stiff action. A tiny bit less still, and the plectrum slips by scarcely activating the string at all. Somewhere in between, we hope, is a sweet spot, but unless you’ve produced a finely made instrument with very close tolerances, that sweet spot may be elusive.  As a practical matter for the builder, this suggests a couple of things. First, to the extent that you can make things precisely yet easily adjustable, that will greatly help in getting things right, especially in the prototyping stages. Second, the parts of the keying mechanism must be steady and stable enough that nothing is prone to wobble or wiggle out of position even by a tiny amount, since tiny misadjustments have big effects.  A third consideration is that the degree of rigidity in the plectrum is crucial. In this connection I have had the nagging thought that my brilliant idea of using the thin spring steel band for this purpose may not have been as brilliant as I thought. And a factor in all of these concerns is the fact that whatever sort of keying and plucking mechanism you create, it must fit in a very small space.  The lateral spacing per key in a standard piano keyboard (when you average out the spacing of black and whites keys) is just over a half inch.  That’s not a lot of room for all that needs to fit in there. A high degree of precision is required to get things right on such a scale.


Don’t get me wrong; I’m glad I made the considerable effort to make this instrument; I do enjoy the way my Minimal Clavier sounds and I expect I’ll use it in various musical contexts going forward. But it must be acknowledged that as a serious keyboard in a world full of admirably made instruments, this one doesn’t really pass muster. I’d still very much like to come up with a more effective design for a simple, home-buildable keyboard string instrument.  If I don’t make that happen, then I hope someone else will – maybe you.

Share This