Introductory materials ( Start here!)

• Introduction only (not collected) Intro to Mathcad The first time we are in the laboratory, please complete this Introduction to Mathcad project.
  (do not turn this project in)
  This introduction will help orient you to the tools that will be used.
•  Review only (not collected) Review Sampling, Aliasing, Quantization, DTFT Mathcad file Sampling, Aliasing, Quantization, DTFT
  pdf file
  Sampling, Quant Noise, DTFT, DFT, and Unit Circle Summary
  CDF, DTFT, DFT, convolution Summary
•  Review only (not collected) Review DFT, Matrix Forms, z-Transform Mathcad file DFT, Matrix Forms, z-Transform
   pdf file
  DTFT and DFT and Matrix Forms Summary
  Difference Equation Summary
  Z-transform Summary
• A quick intro to K64F and mbed with "blinky": see http://thomasweldon.com/tpw/courses/embeddsp/p01embdsp.html

Projects

• Make sure that you complete the introductory materials before you start Project 1.
• Beware: changes are being made to the projects/materials below during the course of Fall 2017
  
• Projects should typically begin during week 2, and approx. weekly thereafter

• Project Mathcad file IIR filters: Bilinear transform, impulse invariance, and window
  Project pdf file
  New Filter Design Summary
  Old Filter Design Summary
  Old Discrete-Time Filtering Overview
• Project Mathcad file Signal/system modeling and Pade approximation
  Project pdf file
  Signal/system modeling and Pade approximation summary
• Project Mathcad file Least squares, left psuedo-inverse, and Prony approximation
  Project pdf file
  Least squares, psuedo-inverse, and Prony summary
• Project Mathcad file All-Pole Models
  Project pdf file
  All-pole models summary
  Random Process ARMA models summary (Yule-Walker)
• Project Mathcad file digital impedance and state variable
  Project pdf file
  See previous project handout on Filter Design plus digital non-Foster circuit handout
  All-pass, invertible, and minimum phase systems
  Notes on poles in magnitude-squared of H(z) and H(s)
• Project Mathcad file Spectral Estimation
  Project pdf file
  New Spectral Estimation
Old Handout Spectral Estimation

• Exam
  Exam Review Handout 
  2006 Exam

• Project Mathcad file Wiener Filters
  Project pdf file
  Wiener Filters
  
• Project Mathcad file Recursive least squares (RLS) adaptive filters
  Project pdf file
  Recursive least squares (RLS)
• Classifiers Project Mathcad file Classifiers
  Classifiers Project pdf file
  

• A quick intro to K64F and mbed with "blinky": see http://thomasweldon.com/tpw/courses/embeddsp/p01embdsp.html
  
• Tutorial for FRDM-K64F and NetBeans:
• Tutorial for project: NetBeans and FRDM-K64F Tutotial and downloads
• Report and demo due for tutorial: due TBD 4/19
  
• Tasks to be included in tutorial report and demo
  
• Measure 1024 points of sawtooth
• Compute FFT of 1024 point sawtooth on FRDM-K64F
• Plot FFT in dB
• Make a table with theoretical and measured sawtooth dc, first, second, third harmonic in dB
• Use Report Template
  
• Project constraints
• Required Hardware: NXP FRDM-K64F plus materials/cables/etc.
                     see https://developer.mbed.org/platforms/FRDM-K64F/
• Required Software: mbed.org developer account to program the FRDM-K64F
                     see https://developer.mbed.org
• Project goals
• Proposals for related exceptional projects may be considered
• All projects must have proposals due TBD 4/21 describing theory, and objectives using  Report Template (3 page maximum)
  
• Choose one goal:
• Design an  audio spectrum analyzer with netbeans interface
• 8192-sample long at 8000 samples/second
  
• All FFT/signal-processing MUST be done on FRDM-K64F using mbed (all calculations must be on FRDM-K64F and NOT in NetBeans)
  
• Demonstrate using iqmidint.au file, played through audio jack of laptop, digitized by ADC on FRDM-K64F
  
• Spectrum displayed in netbeans on PC/laptop, MUST be displayed in dB, and dB MUST be calculated on FRDM-K64F
• You must use a spectral estimation method
• Just an FFT of data is NOT ACCEPTABLE
  
• Explain observed noise floor compared to theoretical floor
• Explain observed signal spectrum compared to theoretical
• Design an audio spectrum analyzer with lcd display interface
• Same specs as above
• Except display spectrum on LCD display
• Examples: see adafruit arduino lcd or similar displays
  
• Final report
•  due during final day of class 
  
• You must email final report pdf plus main.cpp mbed code, and 1 hardcopy due
  
• Format: 4 page minimum, 5 page maximum using  Report Template
• Mbed main.cpp MUST be included as an appendix to final report (does NOT count toward page count of report)
  
• Must have higher quality images
  
• MUST be suitable quality for IEEE publication
• You MUST make full use of all required number of pages, empty space will be penalized
  
• A well-written report/paper is EXPECTED
  
• STRONGLY RECOMMEND that you read IEEE authorship series: How to Write for Technical Periodicals & Conferences
• Clearly describe everything, including:
• describe all variables in block diagrams
• describe your algorithm and flowchart
  
• describe ALL variables in formulas
• describe units of variables KHz, pF, nH, m, s,
  
• describe and clearly label all traces on oscilloscope plots
• describe and clearly label all curves on plots
• describe and clearly label all results in any tables
• Final reports must include as a minimum:
• Introduction summarizing IEEE references in bibliography (NO WEBPAGES for references!)
  
• Theory for your spectrum estimation method
• Theoretical noise floor calculation for your system
  
• A block diagram of whole system from laptop/netbeans through microrcontroller showing DSP block and ADC/DAC
• Flowchart of software
• Oscilloscope trace or netbeans display showing digitized ADC time-domain input signal IN VOLTS
• Include a "zoom-in" time-domain figure too, if details are not clear
  
• An image/snapshot of your output spectrum plot correctly scaled for Volts-squared in dB
  
• A small table showing parameters below
• Column 1: Parameters: bandwidth of frequency bins, noise floor in volts, spectral noise floor in dB, and one other key measured value
  
• Column 2: Theoretical: show theoretical values for all
• Column 3: Measured: show measured values for all
• A bibliography with 4 relevant IEEE references (no references to weldon, no websites)
  
• An appendix (doesnt count toward page count) of main.cpp mbed codel isting
• Final presentations: 
  
•   Default deadline/due-date: final class period, or by appointment
  
• Minimum slides:
• Title slide
• Overview of talk slide
• Block diagram slide: block diagram of system
  
• Theory slide: showing spectral method chosen and noise floor calculations
• Algorithm slide
  
• Time domain slide: showing netbeans time domain signal digitized by ADC
• Zoom-in time domain slide, if needed
• Frequency domain slide: showing netbeans freq domain in dB
  
• Table: showing important parameters, comparing theory and measured, including noise floor
• Any other important details
• Conclusion slide
• Finally: live demonstration of the final system