1 Musical Instrument of Multiple Methods of Excitation (MIMME) Design Team John Cavacas, Kathryn Jinks Greg Meyer, Daniel Trostli Design Advisor Prof. Andrew Gouldstone Abstract The objective of this capstone design project is to bring the versatility and expressiveness found in electronic synthesizers to an acoustic instrument. This instrument will be able to produce complex acoustic spectral envelopes that cannot be reproduced electronically. The design employs three different excitation methods to mechanically produce vibrations in a string. The three excitation methods are similar to a violin, piano, and guitar because they bow, hammer, and pluck the string. The pitch, timbre, and spectral envelope of the vibrations are controlled by dynamically changing the parameters of the system while it is playing. The overall purpose of this project is to create an acoustic sound design tool that allows the user to implement musical creativity. This product will appeal to composers and producers in the music industry. Design of this product involves evaluating string material, calculating torque requirements, optimizing moments of inertia, and designing a control system. Other steps in the design development involve selecting motors, a solenoid, and a linear actuator. Movable bridge Bow Pick Hammer Motor Resonance Box
2 The Need for Project The overall design goal is to create a novel musical instrument that brings some of the features of synthesizers to an acoustic waveform. Despite the flexibility that electronic instruments provide, many modern musicians and producers prefer the sound of acoustic instruments or analog synthesizers in comparison to newer electronic synthesizers that utilize DSP (digital signal processor) chips. The common argument many modern musicians contend over is whether the reliability and convenience that DSP powered instruments provide is worth losing the warmth that physical instruments supply. Using a computer chip to generate sound essentially generates the same sound over and over. Even if the sound changes dynamically with time, the sound has been programmed to follow certain rules and the DSP chip will follow them exactly. It has so far proven impossible to program an algorithm that is indistinguishable from its physical counterpart by a trained musician. This is because there are many variables in physical instruments that affect tone and timbre. For example, in a piano one string vibrates from being hammered, but other strings that share harmonics also vibrate sympathetically. In addition, MIMME also differentiates itself from all other past or similar projects by approaching the design of a musical robot from a fresh perspective. While other projects have attempted to replicate the playing of one particular instrument, MIMME is inspired by the sound design capabilities of electronic synthesizers. By bringing the nuances of physical strings and different excitation mechanisms coupled with features such as a dynamic damping system and automatic vibrato, MIMME produces a genuine acoustic sound. MIMME has the capability to adjust several parameters dynamically, allowing its timbre to change over time. Rep 3.8. The Design Project Objectives and Requirements Functionality was the main focus for design of each mechanism. Design Objectives The objective of this capstone design project is to bring some of the versatility found in electronic synthesizers to a physical instrument. To accomplish this objective the following design goals must be successfully completed. 1. Play at least 3 distinct timbres This is analogous to waveform selection in the oscillators found in electronic synthesizers. It is the starting point of a sound s timbre. This is accomplished by using at least 3 different excitation mechanisms that determine the shape of the vibration on the string. Rep 6. 2. Adjust String Length Adjusting the string length changes the fundamental frequency dynamically, which changes the tone or pitch of the sound being produced. A motorcontrolled movable bridge accomplishes this. Rep 4.5 3. Produce Super Human Rhythm Super human rhythm allows for a deeper sound design by allowing the MIMME to do things human musicians cannot do. This can be obtained with the speed and program capabilities of the string
3 Design Concepts considered A minimum of two design options must be investigated for each excitation mechanism. The final design for each mechanism is selected based on optimal functionality. Recommended Design Concept Each of the design options proposed were considered. The final designs were selected based excitation mechanisms. The design goals listed above are ranked by importance. To accomplish these goals, the design of each mechanism needs to be optimized to allow for quick excitation without creating large motor and controller requirements. A controller system must also be designed to control the excitation of each mechanism and to allow for real-time control. Design Requirements Functionality is the main focus for design of each mechanism. Optimal functionality can be obtained by minimizing moment of inertia and torque required, which reduces the mechanism size and motor requirements. Speed and control of the actuation mechanisms is also critical in obtaining superhuman rhythm. Negative system responses such as overshoot need to be prevented. Multiple designs must be considered for each facet of the device. pairwise comparisons are the main decision tool. String material, bridge, excitation mechanisms and actuation mechanisms must be selected. Pairwise comparisons are used to make the selections. Rep 4.2-4.5. For each excitation mechanism a minimum of two designs must be fully investigated. These design concepts are discussed in detail below. Two initial designs are available for the plucking mechanism. In the first design, several guitar picks are attached to an aluminum stage, which is controlled by a stepper motor. The guitar picks are clamped in place between two low stiffness materials to reduce the elastic modulus of the picking system, increasing its flexibility, and preventing the string and the pick from breaking. The second design has a smaller plastic stage, which reduces the torque required to pluck a string, as well as the moment of inertia. Rep 4.1.1 The first design for the hammering mechanism is very similar to the design of a piano hammer. A stepper motor and ratchet gear actuate the hammer, and a torsion spring prevents the hammer from remaining in contact with the string, minimizing damping. The second hammer design utilizes a linear solenoid to actuate the hammer into striking the string. The hammer then recoils from the force of the first vibration of the string and the force created by a spring on the opposite end of the hammer arm. Rep 4.1.2 and 6.3. For the first bowing design, a motor controlled wheel excites the string by rubbing against it, similar to that of a hurdy-gurdy. The wheel has a small cutout in it to allow for it to rotate in and out of contact with the string. When the wheel is brought out of contact with the string, it is not damped, allowing other excitation mechanisms to be activated. The second design uses a solid wheel, with no cutout. This version still utilizes the fundamentals of a hurdy-gurdy to excite the string, however the entire hurdy-gurdy wheel is moved in and out of contact with the string by a variable position linear actuator. Rep 4.1.3 and 6.1 1. Design Description Of the designs discussed previously, the following are chosen based on functionality and design optimization.
4 on functionality and design optimization. For the final design of the bowing mechanism, the smaller mechanism is selected. This version uses a small nylon wheel controlled by a stepper motor. The wheel is brought in and out of contact with the string with a variable position linear actuator. The use of a solid wheel, without a cut out, allows the string to be bowed smoothly for an infinite amount of time. This increases the creative flexibility for the end user. The small design also reduces the size requirement for the entire device. Rep 6.1. Solenoid Spring For the plucking mechanism, the smaller design is selected for the final device. This design has four guitar picks that are attached to a base, which is controlled by a motor. The main difference between the two designs investigated is the size of the base. Since the selected design has a smaller base than the first, it has a smaller moment of inertia, reducing the motor torque and space requirements. The simplistic attachment procedure for the picks also increases the user-friendliness of the device. Due to the short lifetime of picks, the need for an exchange is foreseeable. Rep 6.2. The final hammer design is the linear solenoid design. This design uses a linear solenoid to actuate the hammer to strike the string. The hammer then recoils from the force of the first vibration of the string and the force created by a spring on the opposite end of the hammer arm. The solenoid design is selected for its ability to quickly engage and disengage the string. The other design lacks the ability to be quickly disengaged. Due to the influence of the hammer-string contact on sound production, this design feature is determined to be a necessity. Rep 6.3. 2. Analytical Investigations Many analytical investigations are required throughout the design process. These investigations can be found in section 4 of the report. For clarity, only two of the investigations are discussed in this summary. These investigations focus on the selection of a motor and belt for the string length adjustment mechanism. The selection of the belt for the string length adjustment mechanism is important to ensure that the design goals are reached and the cost is minimized. A roller chain, synchronous belt, and V belt are compared by Pugh analysis. Efficiency, noise, speed capability, cost, and precision are some of the criteria used in the evaluation. Rep 4.5. Based on this evaluation a synchronous belt is selected. Further calculations determine the exact requirements of the belt. Based on the Pugh analysis and mathematical modeling a specific synchronous belt is selected. A motor must be selected early in the design process to allow for the development of the control system. Three methods are used for the motor selection process. First, research must be completed to obtain speed capabilities, programming requirements, precision, and the torque available with each type of motor. Stepper, DC, and servo motors are considered. To obtain information not addressed in literature, two controls experts must be contacted. Information such as sound produced and controller requirements are determined. Analytical calculations determine the requirements of each subsystem. Stepper motors are selected based on the results of the evaluations. Rep 4.4.
5 3. Experimental Investigations One of the experimental investigations completed in this design process focus on string selection. The string in the MIMME robot is fundamentally important because it is responsible for sound generation. The characteristics of the string determine the timbre of the sound. A string that has suitable characteristics such as high tensile strength, low inharmonicity, and rich timbre is highly desirable. To select the best string, a physical modeling software, Modalys, is used. It converts string properties such as elastic modulus into sound quality. Seven strings are selected based on the richness of timber predicted by the software. A pairwise comparison of the sounds from each string with each mechanism helps determine which string works best. 4. Key Advantages of Recommended Concept The main advantage of the proposed design is the ability to produce complex acoustic spectral envelopes that cannot be created by an electronic synthesizer. This provides new opportunities for sound design that are not currently available in the music industry. This device has three excitation mechanisms for one single string. This concept is also innovative. The mechanisms are designed so that it is possible to activate one without interference from the other mechanisms. The ability to excite the string by one mechanism or multiple mechanisms at a time is an important advantage. Currently there is no comparable device on the market. Software with a graphical user interface has been developed to provide an easy to use controller. The software is compatible with industry standard MIDI controllers for intuitive musical performance. Financial Issues The proposed design is over 50% cheaper than most electronic synthesizers. The prototype cost for this device is just under $800. Currently most electronic synthesizers cost $500 to $2,000 making this prototype affordable for its desired market. Due to the fact only one prototype was built, cost would decrease if components were purchased in bulk. This would result in a less expensive product when manufactured in greater production numbers. Recommended Improvements A dynamic damping system would significantly improve expressiveness of the instrument. It is recommended that a dynamically variable damping system be implemented so that the volume envelope may be controlled on the fly as is possible in electronic synthesizers. This addition would significantly improve the expressiveness of the instrument. Other improvements include the addition of an automatic whammy bar to create vibrato effects, as well as adding a dynamic aperture control to the opening of the resonance box to simulate tremolo.