The Automatic Guitar Tuner AGT Project Oral Presentation

1
The Automatic Guitar Tuner (AGT) Project Oral
Presentation

• Presented by
• Paul Bradley
• Cody Campbell
• Paul Schwartz
• Matt Whitehead

2
Motivation for the AGT

• Tuning an instrument to satisfaction can be
  both time-consuming and difficult.
• Once the desired tuning has been achieved,
  there is no guarantee that the tuning will remain
  intact for any given amount of time.
• Varying temperature
• Varying humidity
• Aggressive playing styles
• Accomplished musicians find it necessary to
  retune their instruments several times per hour
  during moderate use.
• Due to the high cultural popularity of the
  guitar, an automatic guitar tuner would be
  greatly appreciated by the guitar musician
  community as a whole.

3
Market Research

• Everyone we have talked to has had only
  positive responses to the concept of an automatic
  guitar tuner.
• Currently, there are several types of
  commercially available semi-automatic guitar
  tuners available for purchase.
• Merely display the offset direction of the
  pitch being played in comparison to the desired
  pitch.
• Some also give the musician an idea of how far
  off the instruments pitch is from the desired
  pitch.
• There does not appear to be a commercially
  available device which completes the manual
  tuning process automatically with minimal user
  intervention.
• Observation was made by browsing merchandise in
  music stores and reading music hobbyist
  magazines.

4
Market Research (cont.)

• There are a few patents for automatic guitar
  tuners.
• One was designed by a student in England (2001)
  and was found through internet searching.
• U.S. Patent 6,184,452
• Another is from 1977, so the AGT concept is not
  new.
• U.S. Patent 4,018,124

5
Objectives

• Objectives are broken down into three levels of
  achievement.
• Level 1 represents the minimum required
  objectives that would result in the realization
  of a successful design project.
• Levels 2 and 3 represent additional objectives
  that may be met after the Level 1 objectives have
  been fulfilled.
• The initial implementation of the project may
  be mounted separate from a guitar, such as on a
  wooden platform.
• If time allows, the AGT will be physically
  mounted onto a guitar.

6
Level 1 Objectives Specifications

• Tune a guitar string to within a 10 centitone
  frequency tolerance.
• Corresponds to about /- 0.58 of the nominal
  frequency
• (/- 0.5 Hz) for the lowest guitar string (Low E
  String).
• Require less than 6 seconds to successfully
  tune the string.
• An adult with minimal musical training should
  be able to use the AGT easily after reading a
  brief set of instructions.

7
Level 2 Objectives Specifications

• Tune all six guitar strings, one at a time, to
  within a 10 centitone frequency tolerance for
  each string.
• Total weight of the guitar mounted portion
  should be kept under 10 pounds.
• The AGT must not impede the playing area of the
  guitar.
• The AGT should not make holding the guitar
  uncomfortable.
• Power consumption while tuning each string
  should be kept under 6 watts.

8
Level 3 Objectives Specifications

• Tune all six guitars strings simultaneously to
  within a 10 centitone frequency tolerance for
  each string.
• Design the AGT to be battery powered and ensure
  than it is capable of running for at least 100
  tunings per battery pack.
• Robust against user misuse
• Compatibility with over 50 of guitars (both
  acoustic and electric)
• Consumer cost should be under 500

9
Top-Level Block Diagram

10
Second-Level Block Diagram

11
Control Unit
Central Control Unit
interface control signals
interface control signals
6
desired frequencies
6
Signal Processor
Motor Control
raw pickup signal
6
6
measured frequency
motor actuation (PWM)
12
Signal Processor

• Two approaches. We are favoring the first.
• 6-way microcontroller controlled analog filtering
• LMF100 switched capacitor filters and a frequency
  to voltage converter, all driven by a PIC 12F675
• Single Microcontroller using DSP techniques.
• Motoola 555 at 40 MHz with hardware floating
  point doing a 1024 point FFT

13
Central Control Unit

• More of a conceptual system block than a real
  component.
• Comprises the master software that takes the
  input from the user interface and runs the show.
• The system will idle in a low-power standby,
  with the filters off and microcontrollers
  possibly in a wait state.
• Turns things on and begins the tuning process,
  monitoring the results and signalling the user
  when things are complete.
• Could be implemented on a separate processor or
  in separate compute time on any that has time to
  spare.

14
Motor Control

• Microprocessor controlled PWM. Input from Signal
  Processor as digital freqency measurement, output
  to Motor System as logic level PWM.
• An 8-pin PIC can make a 19 kHz PWM signal from 0
  to 100 duty cycle and have time for a little
  math to spare.
• Motorola 555 has a 32 channel programmable timer
  unit capable of producing PWM.
• A suitable algorithm has yet to be determined.
  Certainly we can approximate classical feedback
  in the digital system, as well as a simple
  up/down control to get close to the desired range.

15
Motor System
Power Electronics
6
6
6
Motors
motor actuation (logic level PWM)
string frequencies
motor actuation (drive level PWM)
16
Power Electronics

• H-bridge driver
• Bypass diodes
• MOSFETs compatiblewith logic signals
• Alternately the whole system is available on
  achip (LMD18200 3A)

reverse
Vcc
forward
M
17
Motors

• We plan on winding our own, of course.

18
Motors

• Driving the string through a tuning peg.
• Brushed DC servo motors, for torque, size, and
  simple control (PWM)
• Motors will need a planetary gearhead to operate
  at the low speeds and high torques required.
• Many to choose from Maxon, Douglas
  International, MicroMo (Faulhaber), Micro-Drives.
• To turn a peg we need about 30 oz-in or 0.2 Nm
  at about 120 rpm. We can get that in a 1 inch
  diameter package.

19
Pickup
20
Pickup

• The pickup block detects the frequencies of the
  guitar strings and sends a signal(s) representing
  the string frequencies to the signal processing
  unit
• Two alternatives
• Use a divided guitar pickup that senses each
  string frequency individually and transmits six
  separate signals to the signal processing unit
• Use a standard guitar pickup that sends one mixed
  signal that represents the frequencies of all six
  strings to the signal processing unit

21
The First Alternative

• This alternative requires the utilization of a
  divided pickup
• More expensive (Approximately 230 for the Roland
  GK-2AH compared to 40 for a standard pickup)
• To be effective, this pickup must greatly reduce
  the amount of crosstalk between signals
• This will minimize the amount of filtering and
  signal conditioning
• If it does not reduce crosstalk, it will just be
  an expensive standard pickup
• Outputs six individual signals representing the
  frequencies of each string

22
The Second Alternative

• Use a standard guitar pickup to output one mixed
  signal representing all six string frequencies
• If this approach is used, a significant amount of
  filtering and signal processing will be required
• This will add a great deal of complexity and cost
  to the AGT
• DSP boards cost a lot of money
• One design goal is to avoid having a computer in
  the system
• Thus, the first alternative is preferred

23
To Use the Divided Pickup

• Several key issues must be researched
• How to interface with the pickup
• Analyze the signals that the pickup outputs
• Current levels, etc.
• Observe the amount of crosstalk
• Is it acceptable?
• How much filtering is required?

24
Pickup Conclusions

• To summarize
• The divided pickup will enable the AGT to tune
  all six string simultaneously
• May lower cost of system since it will
• Reduce the amount of filtering
• Erase the need for an expensive DSP or personal
  computer
• The Roland GK-2AH is compatible with both
  acoustic and electric guitars which was an
  original design goal of the AGT

25
User Interface
26
User Interface

• The user interface will allow the user to
  communicate with the AGT
• It will include
• Activation button or switch
• Begin Tuning button
• Cancel Tuning button
• LEDs that signify the current status of the AGT
• Any other item that is found to be necessary for
  the user to communicate with the AGT

27
User Interface
ppmc enterprises
AGT4000
tuning
done
O.O.T.
tune
cancel
28
User Interface

• User interface will send signals directly to/from
  the Central Control Unit (CCU)
• The CCU will process the requests and send the
  appropriate signals back to the user interface
• To summarize, the user interface will
• Make the system easy to use
• Allow the user to start or stop the tuning
  process
• Inform the user the exact state of the AGT at all
  times
• Be inexpensive and have low power consumption

29
Power Supply

• Job of power supply block is to provide the
  correct amount of power to each of the other
  blocks
• Two possible options
• Accept power from external source
• e.g. standard 120V, 60 Hz power outlet
• Contain internal power source
• e.g. batteries
• Could include both options

30
Power Supply

• Accepts either one or two inputs
• Signal representing whether the system should be
  on or off
• Possibly the input power signal from the external
  source
• Power supply will convert the internal or
  external power source into the appropriate levels
  and distribute power to the other blocks

31
Math AnalysisString Mechanics

• Main formula T (2Lf)2µ
• T tension N
• L length of guitar (fret board to bridge) m
• f string frequency Hz
• µ mass per unit length of each guitar string
  N/m

• Example calculation for the low E string
• T (2Lf)2µ 2 26.75 inches (2.54 cm / 1
  in) (1 m / 100 cm) 82.407 Hz2 (0.0098 kg
  / meter)
• T 122.6611 N
• Excel theoretical values 10x smaller
• Different Guitar
• Friction

32
Math AnalysisMatlab Work

• T (2Lf)2µ
• Second order formula
• First order and second order approximations using
  least-squares curve fitting
• Max first order approx. error

33
Estimated Budget

• Low Estimate
• Single Microcontroller
• High Estimate
• Multiple or more powerful Microcontrollers
• Higher Quality Sensing Device

34
Schedule
35
Conclusions

• Motivation
• Market
• Objectives
• Main goal
• Subsystems
• Math Analysis
• Budget
• Schedule

• Tuning is difficult and time-consuming
• Product doesnt exist
• Three levels of success
• Simultaneously tune all six strings
• User Interface, Pickup, Signal Processor, Control
  Unit, Motor System, Power Unit
• String Mechanics
• 600 to 2000
• Design Signal Processor, Motor Controller, User
  Interface, Mounting Device

36
Any Questions?
W