MTE3013 - Designing Musical Interactions

O.M.N.I.
Opto-coupled Modular Noise Instrument


Concept

The thoughts and ideas behind O.M.N.I. came from a 'hacking' workshop run by John Bowers in the Sonic Arts Research Centre (SARC) at The Queen's University of Belfast. At this workshop, I was introduced to simple sound producing circuits, as well as the concept of opto-coupling:

Opto-coupling is mainly used in technology where high voltage components communicate with low voltage components, and in order to protect the lower voltage components, light is used as the medium of transfer, avoiding a potentially damaging physical connection.

I was intrigued by the idea of having circuits interact without having physical connections, and it was this idea I wanted to explore, by creating an instrument in modules that didn't have physical connections between those modules that produced sound and those that were affecting their output.

In this instrument, opto-coupling is used to influence the output of sound circuits in interesting ways.

There are four modules that make up O.M.N.I., which are listed below:

- Module 1: Light Module
- Module 2: Sound Module I
- Module 3: Sound Module II
- Module 4: Amplification Module


Initial Design

The initial design for the instrument was to build a set of four boxes: two sound producing, one light producing and one amplifier.
The light circuit would have two LEDs, one on each side, that flashed at regular intervals controllable by the performer. These would then be read by LDRs in the sound modules to produce changes in the output.
The sound modules would consist of two CMOS 4093 based square wave oscillators, one of which being frequency modulated. The opto-coupling would perform amplitude modulations on the sound modules making them produce output when receiving light. The layout of the modules as the performer would see is outlined below:

Early tests of the instrument were built on 'breadboards' and proved incredibly useful as it offered the ability to quickly swap components to get different results. Although the final performance layout of the modules could not be tested, it was evident that the opto-coupling would produce interesting changes in the way that the sound modules behaved. After these tests, it was clear that the sounds from the two CMOS 4093 circuits were very similar and that a change to one of the sound modules was necessary.


Developed Design

The initial prototypes of the instrument were expanded, and after some research into other instruments being built in SARC, I came across the Atari Noise Board (ANB). It was designed by Cavan Fyans and is based on the Atari Punk Console circuit created by Forrest Mims. It produces a wide range of sounds and has its own unique 'character' as a synthesiser, which is similar enough to complement the other sound module and yet still different enough to be interesting overall. As this is supplied as a circuit board, it immediately lends itself to alterations and modifications.

The design of the overall instrument changed as a result of the inclusion of this board, given that it does not require amplification, and so the amplification circuit was removed during testing. The modifications made to the ANB were temporary at this stage, as the results were unknown, and this proved vital.
In order to adapt the ANB for use with the opto-coupling concept, I connected an L.D.R. through the 'positive rail' of the power supply, reducing the power going to the device dependant on light received. The modifications to the board resulted in severely reduced output given that it was being 'starved' of power. This was easily solved by the re-addition of the amplification module to the instrument. However, a more serious problem emerged, reduced musical output. 'Power starving' the board, restricting the power supply, meant that the integrated circuit board that controlled the timing, and therefore all of the output, wasn't functioning correctly.
The idea behind modifying the circuit was to increase the musical output, not restrict it, and so it was necessary to redesign the modifications completely.
The ANB has two controllable parameters, outputting an arpeggio with one control altering the note in the arpeggio and the other controlling the range of the arpeggio. I began altering the circuit by replacing these controls with LDRs and found that the best parameter to control using the opto-coupling was the arpeggio range. However, even with the removal of the 'power starving' technique, the output of the circuit is still severely reduced, and so the fourth amplification module was definitively re-introduced.


Final Design

Sound Module I

The CMOS 4093 based frequency modulated square wave has two controllable parameters: the LFO speed and the main pitch. I had thought about having pitch as the parameter altered by the opto-coupling LDR, but decided it would be more interesting to control the LFO instead as it gave more sonic variation during prototyping and gave a greater contrast with the second sound module. The LDR for this is placed inside housing instead of flush mounted to enable shielding from extraneous light. The other parameter is controlled using a potentiometer attached to the top of the housing for the module. The final feature of this module is a momentary on/off switch which enables the performer to easily create more complex rhythmic patterns than those created only using the LFO speed control. To connect with the amplification module, or an external speaker system, this module has an RCA out connection and I have purpose built cables for use with these.

Sound Module II

The ANB again has two parameters. The arpeggio range is controlled by opto-coupling, as in the previous design stage, as it produces very interesting modulations when used with this and as with other sound module, the LDR is placed inside the housing to shield from extraneous light.The other parameter of which note is being played in the arpeggio is controlled by a potentiometer placed on housing. The final feature of this module is a PTM/PTB switch, enabling the module to be switched on and left running until the switch is pressed again to turn off. This proved more interesting during prototyping given the types of sounds produced by this module. In addition to this, as with the other sound module, this module has an RCA out connection to connect to either the amplification module or an external speaker system.

Light Module

This module consists of an oscillator controlling the rate at which two LEDs blink. The potentiometer and capacitor values where chosen to give ranges from a slow blink to constantly being on, as this produced a wide range of effects in the sound module. Both LEDs blink at the same time however the sound modules don't necessarily react in the same way to same stimulus. One LED is mounted on each side of the housing, as well as the potentiometer mounted on the top for ease of control.

Amplification Module

The amplification module consists of an LM386 based amplification circuit, modified to have two separate RCA inputs which are adjustable with potentiometers mounted inside the housing. This is connected to a small speaker, again mounted inside the housing. During prototyping, I discovered that if I manipulated the speaker cone whilst it was in use, this provided a filtering effect, and so I mounted all of the circuitry 'upside-down' in the housing to allow the performer to do this whilst using the instrument.


Conclusions

I am incredibly pleased with O.M.N.I., as it has improved greatly since the initial ideas. There are several subtle nuances with the instrument that make it enjoyable to play, most notably with the parameters set at just the right position, it is possible to obtain two separate rhythms from each sound module and have each of these being affected completely differently by the light module, almost creating a set of four intricate rhythms. The level of unpredictability that is introduced with nuances such as this make performing with the instrument an interesting challenge and it is never the same twice.
The positioning of each component on the housing was thought out in great detail to ensure an instrument that was easily playable, and I am most pleased with the positioning of the momentary switch for the CMOS 4093 module, resting just under the palm of the hand whilst holding the potentiometer enabling the performer to lightly press down in a natural position whilst controlling the other parameters, and the fact that the instrument is modular enables infinite expandability, and I plan to continue to design extra modules for use with this instrument in the future.

Having said this, there are still some improvements that could be made, mainly to the amplification module, but the modularity of the instrument makes these incredibly easy to integrate. The amplification module contains a very simple mixing circuit to combine the two signals it receives, and being very simple it introduces its own effects to the output of the instrument, mainly one signal drowning the other and performing its own modulations based on this. This may be viewed as an interesting nuance from the instrument, but the amplification module was only ever intended to amplify the signals, not interfere with them. This could be solved by using a more robust mixing circuit design, or the integration of a separate amplification circuit and speaker for each signal into the module.