Have you ever thought about why the keys are arranged the way they are on a piano? The piano keyboard is extremely well established and few people question if its a good or bad keyboard layout. It has taken quite some time for me to realize that the piano keyboard really couldn't be an ingenious device. From a historical viewpoint the key arrangement on the piano is actually a remodification of some really old designs that were used in the renaissance period. The Janko keyboard is named after its inventor Paul von Janko which was a hungarian matematician. Indeed I could spend all day and night just talking about the theory and background of the piano keyboard versus the so called Janko keyboard and I believe this wikipedia article explains it pretty good.
I like playing synth keyboard although I've never been able to play in various keys in order to play along to any music without transposing. Now transpose can sometimes like on my portatone require menu tweaking in order to adjust it. This made me think of an instrument that could have its own transpose keyboard to quick adjust the notes on the piano. But then I discovered that the Janko keyboard concept exists even though practically no company make them except a Japanese company called Chromatone. This is of course a nice device but with its $1500 price and $300 shipping cost its well outside my budget I'm afraid. The reason it costs ten times a regular portatone is that it is some kind of prototype and those who order it might have to wait a year to get it. One more thing is that I want a two manual organ and needless to say that it would get terribly expensive. As a fanatic Janko supporter I've got no other option that to take matters in my own hands. As a matter of fact I've practiced on an emulated Janko consisting of two ps/2 keyboards and a microcontroller. This was an earlier project to evaluate Janko and to practice. I just loved the playing experience. Now I need the real thing.
The keys are the Cherry MX series bought from DigiKey. I've mounted them on a rack consisting of small phenolic board pieces. Brass
spacers are used to hold them in position. The wires are soldered underneath the boards and are connected to an IDC connector terminal.
The long IDC cables are visible behind the keyboard modules.
I've managed to make keycaps from polycarbonate plastic sheet.
Here's a YouTube demonstration where I play the thing.
I've already developed a design with Atmel AVR microcontrollers. I started with the electronics part before I had any idea what pushbuttons to use. Right now my code responds to crossover contacts and is touch sensitive. But since the Cherry switches don't support velocity sensitivity I had to remake my code to fit these new circumstances. The AVR's communicate via SPI (don't confuse with ISP) and I'm using a 'raise hand' handshaking protocol. The other AVR's besides the main unit are all gonna be SPI slaves and the main unit the SPI master. Now since the slaves might suddenly have data that they want to transmit (due to keystrike) they will gently raise their hands in order to notify the master that they've got information. The master is then supposed to watch out for these hands in order to collect the data. This approach would also guarantee that no data collitions would occur since the master will only attempt to read from one slave at the time. The slave will hold their buffers until master collects them one by one.
To the left is the keyscanner board. The events are then sent serially through the SPI interface to the main unit that collects
all commands from all units.
This is the main unit. The hardware for MIDI input is implemented in case I want to make settings adjustable from softsynths.
Solder side view. To the left is the keyscanner and to the right the main board. The approach is clearly oldschool. A modern
implementation would probably have one tiny microcontroller on each subboard and serial communication.
The Janko keyboard is one member of the family of so called isomorphic keyboards. This means that all intervals and chords with a certain shape will always sound the same but at different pitch according to its position. This is an extremely advantageous property of a musical instrument. Not only is it sensible but also very practical. Now lets take a look at a generic isomorph instrument with hexagonal pattern. The variables A and B represent integers for semitones in the direction facing away from the sides of the hexagon.
On the Janko the value of A and B is simply A=B=1. This will consequently lead to that +(A+B) will become 2 which is the number of semitones for a whole tone or a full note. This will also lead to the feature that all notes located on the left side will always be lower and all keys to the right will be higher pitched. This property coincides with a regular piano. Other instruments exist that also shares the hexagonal lattice layout for notes but they will have other value for the variables.
The following table mentions a couple of examples for isomorphic solutions for isomorphic instruments with hexagonal lattice.
| Keyboard Type | A | B | A+B | Additional Information |
| Janko Keyboard | 1 (one semitone) | 1 (one semitone) | 2 (one wholetone) | |
| Button Accordion (C System) (most common) | 1 (one semitone) | 2 (one wholetone) | 3 (minor third) | |
| Button Accordion (B System) ("dugmetara") | 2 (one wholetone) | 1 (one semitone) | 3 (minor third) | |
| Wicki/Haydn | 7 (fifth) | -5 (fourth) | 2 (one wholetone) | |
| C-Thru Music Axis (Harmonic Table) | 3 (minor third) | 4 (major third) | 7 (fifth) | Image Rotated 90 degrees CCW |
| Array Mbira | 5 (fourth) | 7 (fifth) | 12 (octave) | Image Rotated 90 degrees CCW |
Wholetone Basics at www.thecipher.com
recommended reading!
Wikipedia Article about Janko
Japanese Chromatone Project