The next several weeks in lab will be primarily devoted to the design and development of a project related to signals and systems. You are free to choose the topic for your project. Some suggested topics are listed on this page, and some web links are provided. You will have the opportunity to give an oral presentation/demonstration of your project on November 28 and 29, and you will submit a written report.

For October 24 and 25: Please list some potential project ideas, and identify your lab partners for the project. Write a draft abstract describing your project before you leave lab on October 24 and 25. Also, please finish items from Lab 6, if necessary. Be sure you can use the FFT on the oscilloscope, and that you know how to use the Keithley A/D, D/A card from Matlab.

For October 31 and November 1: Please submit an abstract that describes your lab design project. Include the title, the names of the participants, and a brief description of what you plan to do. You may revise your plans as you work over the coming weeks, but do your best to write an abstract based on your current understanding of your project. You should write a draft of the abstract during the lab sessions on October 25 and 30, and refine the draft for October 31 and November 1.

Design Project Overview

You can work on the project individually, or in groups of two or three students. You must work with students in the same lab session. Projects from larger groups should be more ambitious. Be creative, and choose something that interests you!

Your project can be a purely analog system, a digital signal processing system, or a combination of analog and digital. The A/D and D/A cards are available to facilitate data acquisition and real-time implementation of your project. You can sample data with the A/D card and process it in Matlab, for example.

It is also possible for your project to be more of a research paper than an implementation. However, if you choose this type of topic, I would like you to try to implement (or simulate in Matlab) some aspect of the topic.

Digital filtering is one application of digital signal processing. A digital filter has a similar objective as an analog filter, namely to accentuate certain frequencies while attenuating other frequencies. The digital filter is actually a computer program that processes the sampled data. The Keithley cards will allow you to do real-time digital filtering, but at a limited sample rate.

Some Project Ideas

• Analog communication system to transfer analog signals such as speech and music. The system can be implemented with analog components, or the modulation/demodulation can be performed digitally with Matlab. We will discuss amplitude modulation (AM) in class, so you will get an understanding of how this works.
• Digital communication system to transfer bits (0's and 1's). Implement the system in real-time by generating waveforms in Matlab, sending them into a channel (a circuit) with the D/A card, and then sampling the output of the channel with the A/D card. Try to recover the bits.
• Digital speech processing. An example project is to recognize speech and build a device that turns a motor or a light ON and OFF based on voice commands.
• Audio signal processing. Examples include a digital equalizer system, and special effects such as delays, echos, and reverberation. You might also design and build an analog equalizer.
• Telephone touch-tone dialing: generation and detection of DTMF tones. Try to dial a phone, and detect the tones from an actual phone. Investigate alternative approaches to identifying the frequencies that can be more efficient than band-pass filters. (I can provide you with articles on this.)
• Spectral analysis with the fast Fourier transform (FFT). Interesting projects can be considered to detect signals that are buried in noise, or to remove interference signals.
• Time-domain system identification of RC circuit parameters. We have developed an algorithm that determines the time constant of a first-order system from experimental measurements. The algorithm is "recursive", in that it begins with an initial estimate, and then updates that estimate with each new measured data point. For more information, see
  http://www.eg.bucknell.edu/~kozick/elec32004/ParamEst.pdf , or in HTML format at http://www.eg.bucknell.edu/~eg100ee1/paramest/final.html
• Frequency-domain identification of RLC circuit parameters. See the following links for more information: Notes and and MATLAB program lsq.m
  Using this approach, you can estimate the values of R, L, and C in a circuit with as few as three measurements! You can automate the procedure so that the measurements are computer-controlled.
• Encryption/decryption (or scrambling/unscrambling) of speech signals for security.
• Acoustics localization of objects using an array of microphones. Try to estimate the location of the object by processing the sound signals measured at each microphone.
• An electronic guitar tuner (may be done with analog filters, or digitally).
• Digital synthesis of acoustic guitar and other musical instruments.
• Digital image processing, including image deblurring, noise removal, etc.
• Deconvolution: Try to recover the input signal to a system by observing the output of the system. If the impulse (or frequency) response of the system is known, then this is not too hard. Many times the system response is not known, and this leads to "blind" deconvolution.
• An adaptive notch filter to remove sinusoidal interference from a signal. The filter automatically adjusts itself if the frequency of the interference changes.
• Any application of analog and/or digital filters.