Managerial Design Review Presentation

Report
P11227- ACTIVE NOISE CANCELLATION
Kyle Desrosiers (ME)- Team Lead
Brad Fiedler (EE)
Greg Wodzicki (EE)
Chris VanWagenen (EE)
George Slack- Faculty Guide
Acknowledgements:
Cenco Physics
Texas Instruments
PROJECT DESCRIPTION
• This project aims to attenuate sound level output of a given
source using an Active Noise Cancellation system (ANC).
• First of a series of projects ultimately targeting attenuation
of Internal Combustion Engine.
KEY CUSTOMER NEEDS
• Prove concept of sound attenuation via Active
Cancellation
• Demonstrate technical challenges limitations of ANC
• ANC system fully functional to work on a given input
KEY ENGINEERING SPECS
• Noise reduction- Source only vs ANC system on
• Target: >6dB
• Ideal conditions: ~17dB
• Effective Frequency range
• Ideal: 20-20k Hz (Human Auditory Range)
• Actual: 20-1000 Hz
• Processing time
• Ideal: 0.5 ms
• Actual: Unknown
• Speaker Output Level
• ~100dB
CONCEPT SUMMARY
• Sound Waves:
• Propagation of compression and rarefaction through air
Source: http://misclab.umeoce.maine.edu/boss/classes/SMS_491_2003/sound/sound_wave.jpg
SUPER POSITIONING OF WAVES
• In Phase
• Out of Phase
SYSTEM ARCHITECTURE
• Microphones: Input Signal, Error Signal
• Power Amplifier: Behringer EPQ1200
• Speaker: Goldwood GW-210/8
• DSP:
• TAS3108EVM2 Eval Kit
• C5505 DSP Eval Kit
DESIGN SUMMARY
• System Spec’d for Lawn Mower Engine
• Wye Configuration, Dipole Configuration
• Speakers:
• 10in, 110 Watt RMS
• 86.5 dB Sound Level output
• 40-2200 Hz
• Amp
• 2 x 320 Watts at 8 Ohms
DESIGN SUMMARY
• Microphone
• Typical Frequency Response
• Cheap
• For Initial Testing
• DSP
•
TAS3108EVM2 Eval Kit
•
C5505 DSP Eval Kit
DSP
• TAS3108EVM2 Eval Kit
• Sampling Rate: 192 kHz
• Programmable in Assembly or GUI Interface
• 8 Inputs/outputs
• C5505 DSP Eval Kit
• Sampling Rate: 48 kHz
• Programmable in C++
• TI Support
• 2 Inputs/Outputs
ALGORITHM
• Least Mean Squares (LMS)
•   + 1 =   +  ′  ()
• Modeled in Matlab
• Shown to Work
• Unable to successfully implement in first DSP
• Feedback issues using second DSP stalled implementation
TESTING/RESULTS
• Matlab Simulation
TESTING/RESULTS
• Test Setups Conducted
• Straight Inverted Signal, Equal Path Length
• Wye
• Dipole Box
TESTING/RESULTS
• 4in Pipe, Wye
TESTING/RESULTS
• 4in Pipe, Dipole
• Lawn Mower Sound Wave
• 8dB Reduction
TESTING/RESULTS
• 6in, Wye Connection
• Complex Sounds
• Dependent on sound
• Achieved 8dB reduction
TESTING/RESULTS
• Using Microphones and DSP
• Max Reduction achieved: 3dB
• Feedback through mics limited test range
• Testing conducted at low sound levels
SUCCESSES AND FAILURES
• Successes
• Proof of concept
• Geometry, superposition
• Pure tones
• Cancellation of low frequencies with inversion
SUCCESSES AND FAILURES
• Failures
• Programming original DSP
• Programming second DSP, fine tuning
• Microphone Feedback, Clipping
• Unable to implement on non electrical sound source
FUTURE WORK
• Testing
• 6 inch dipole setup
• Back pressure with system in use
• Algorithm debugged
• New microphones
• Eliminate feedback/clipping issue
• System downsized, optimized
• Application to mechanical sound source
SUGGESTIONS
• Computer engineer
• Better knowledge of programming
• Increased budget
• Microphones with proper range, sensitivity, heat
properties
• More expert advice/input

similar documents