(I have recently recieved funding from the Industrial Research Assistance Program of the National Research Council (NRC-IRAP) to work on the following project. As results come in I will post them on the website.)
Development of a True Acoustic/Electric Violin
The violin, as an instrument, has not yet benefited from an integrated approach to amplification and good acoustic performance. With the resurgence of interest of the violin in popular and folk music traditions and the need for amplification of the nuances of a good acoustic sound, there is a demand for an instrument that can perform well acoustically and still fill a hall with a good amplified sound.
Broadly speaking, the player has the option of choosing to attach a so-called “pick-up” to the body or bridge of their instrument or utilizing a microphone. A third option, the electric violin usually has abandoned any acoustic properties and cannot be played without amplification. The electric violin sound is also unsuited to folk or classical music. The intent of this research is to develop an instrument that will perform well in both acoustic and amplified modes.
The technical challenge is to develop a method for the installation of a factory-installed pick-up, probably piezoelectric, which will produce good tone. Violins radiate sound in different modes at different frequencies. At lower frequencies the radiation pattern is omni-directional and large areas of the back and front of the instrument are involved. As the frequency increases smaller and smaller areas of the instrument are involved and the radiation pattern becomes highly directional. In addition, the phase relationships between these areas can vary widely resulting in some cancellation of acoustic output. These phase relationships constitute an important characteristic of what a musician would consider “good tone”.
The advent of PVDF film piezo sensors provide an opportunity to develop an economical wide area sensor to be permanently installed inside a good violin. It is anticipated that a sensor which contacts a significant portion of the violin’s radiating surfaces will develop an electrical signal which is more analogues to the acoustic output of the instrument.
Some of the technical challenges
Piezo materials are available in various materials and thicknesses. The most likely materials of choice are PVDF film and ceramics. With PVDF films the voltage output is directly proportional to thickness at a given strain or stress. Capacitance is also proportional to film thickness. Capacitance directly affects the impedance of the sensor which must be within a set range to match amplifier impedance. The trade off between output voltage, capacitance (i.e. impedance) and sensor thickness must be determined experimentally as the strain imposed by a vibrating violin plate cannot be determined except by test attachment of trial sensors.
Since violins radiate different frequencies from different areas of the violin and the phase relationship between these areas is important, it is unknown at this time what areas of the violin will need sensors. In addition, as PVDF film output voltage is proportional to sensor area, different areas may require different sensor sizing. As with film thickness, capacitance is also a function of film area which will affect impedance matching with amplification systems. The simplistic ideal sensor would be a single wide-area device which would be of the correct impedance and capture a significant portion of the radiated frequencies at a reasonable output voltage. A more realistic system will likely be an array of smaller sensors strategically placed. Testing a range of areas of the vibrating plates at different frequencies should provide data for a preliminary sensor array design.
Attachment of a sensor array to the top and back of violin will have an effect on the radiated sound due to increased mass and possible changes in plate stiffness. It may be possible to reduce such effects by re-designing and re-tuning the violin plates to compensate and retain good acoustic tone.
Musical instrument sensors need to be insensitive to on-stage sound to prevent significant feedback problems. The soundboard of a violin will vibrate in sympathy with loud on-stage sound. It is very important that the ideal sensor array will be highly insensitive to such effects without a significant reduction in desirable output signal.
Very low frequency noise such as bangs and bumps must also not generate a significant signal through the sensor array. The sensor array configuration may possibly be mechanically isolated from such low frequencies and/or the sensor circuit’s cut-off frequency may be selected to eliminate them without a significant loss of desirable signal.
It is possible that the best sensor configuration will be a multiple sensor area. Such an array will require choices to be made regarding how output from the instrument will occur. A single, blended output would be the simplest choice but will require careful circuit design for good tonal balance. Another choice would be to provide multiple outputs; treble, mid-range and bass for example, and allow the player to balance the sound. In either case, the circuit must be designed to produce a good output voltage at the correct matching impedance.