Thursday, 31 March 2011
Casio MT-40 Sequential Voice Switch
What is it? - a circuit added to a Casio MT-40 keyboard to rapidly switch between four voices. It could also be adapted to switch audio or control voltage sources.
How does it work? - I've lost the circuit diagram (I made this back in 2005), but it's pretty basic. The MT-40, in common with many Casiotone keyboards of the early '80s has two keyboard modes; "Play", where the keys function in the standard musical fashion, and; "Set", where the keys are used to select the voice, or Tone as Casio call it. Each white key selects a Tone, the names of which can be seen just above the keys.
The slider switch which toggles between these modes is just out of shot to the right. To the right of that is another switch labelled "Tone Memory", which allows you to change between four pre-selected tones without changing to the "Set" mode. I found that rapidly moving this switch whilst playing made a cool sound, but I could only play one-handed. So I decided to automate the rapid switching.
The circuit I built to do this is a standard sequential switch that I found many variations of online. The larger IC is a 4017 decade counter, triggered by a 555 timer, the speed of which is determined by the value of the capacitor and the potentiometer. Four of the 4017 outputs are connected to the Base pin of each of the transistors (by way of indicator LEDs, and buffered by 1M resistors, as the unit runs at 7.5v). The Collector and Emitter of each transistor are connected to the four pairs of contacts in the "Tone Memory" switch. The 5th output of the 4017 is connected to its reset pin so it counts to 4 rather than 10.
The picture below shows the circuit installed into the MT-40 enclosure. To the far right are the controls - an on/off switch and a rate knob. To the left you may be able to make out the empty slot where the Tone Memory switch used to be.
Further Developments? - I set to building a variation on this which was to have +/- 5v power and transmission gates in the place of the transistors, which would allow me to cycle rapidly through up to 10 audio signal paths (or more if a second 4017 was added). My intention was to build something like a step-wah, but it turned out that somebody was already making one, so it quite literally never got off the drawing board.
Another use that this might be put to is sequentially linking bend-points in a circuit-bent device to create a rhythmic pattern. An external trigger input could be substituted for the 555 so that this could be synched to a multitrack by sending the device an audio pulse e.g. an isolated click.
Credits: Big thanks are due to Imogen on this one!