Introduction

ECE 468 Project Philosophy

"We want to control the position of a motor. We will feedback the position using a potentiometer. What is a ten-turn pot? What voltage should we impress on the pot? Let's compare the pot output voltage to an input reference voltage entered via the key board. Let's A/D the output sensing pot . Let's subtract this pot angle-sensitive voltage from the desired pot voltage entered through the key board. Let's run the motor at a voltage proportional to the error between the desired value and the value measured by the pot, supplied by the D/A. When the motor reaches the desired angle let's turn the motor off."

The above is a bad idea. The PC with data acquisition cards is simply serving as an operational amplifier, a very expensive one!

The ten-turn pot is a good idea for angular position measurement. Suggest that you use +10.0,-10 Volts on the pot. Let the desired position be something that is program generated. Should we, worry about how quickly this is expected to respond, i.e., slew rate, etc? Yes, feedback the pot signal, compare it against desired position to form the position error, then act on this signal with some control algorithm, e.g. a PID algorithm, a lead or lag compensator(s), take the output of the algorithm to the D/A. The D/A won't drive a DC motor. Better use the Kepco op-amp which is good for 5 amps up to 75 volts at many kilohertz. OOPS, but a DC motor is quite nonlinear, with a good bit of stiction. It doesn't begin to turn, even with no mechanical load, until approximately 25% to 40% of rated voltage is applied. Maybe we should use pulse width modulation? In any case we should not turn it off, but control it with a proportional signal.

The above is the beginning of a real solid ECE 468 Project.

"But my interests are digital logic"

These computers have a ?16?-bit parallel TTL interface ports for both input and output. These can be addressed at a ?1 megahertz? rate. This machine could then be used to check out the performance of a digital logic system. Your project could deal essentially only with the digital interface on the PCs, not using A/D's or D/A's. Past projects have done this successfully.

"My interest is communications"

We have extensive signal processing software in the system library. A project could center on developing tools that ECE 330, ECE 430 and ECE 431 might use to establish an Engineering Feel for many concepts that were theoretical mysteries in the course where first encountered.

	The Dell PCs dedicated to this course are fitted with high quality National multi-purpose interface cards which are "keenly" supported by LabVIEW software. This software includes many very special signal processing routines, simulation routines, matrix manipulations, graphics routines, etc.
	An ECE 468 project should not simply use these computer systems as operational amplifiers or as simple switches. Certain "recent" low quality projects used the computers in this manner. An unsatisfactory use of one of these computers would be as follows.

	How about carrying out a variety of demonstrations of digital filtering applied to audio signals?
	It would be interesting to see how bandwidth impacts telephone communication, further how the number of bits in A/D and D/A communication of voice impacts the quality of transmission.
	It is interesting to investigate how a person speaking responds to hearing his spoken words returned after a delay, possibly a different delay to each ear.
	Let's do something in sound synthesis???
	How about looking at two dimensional FFTs applied to graphical data? We have the software.

"I am interested in manufacturing"

In the past we have had a variety of projects that actually manufactured a project, candles, packages of cheese, model air-plane wings (is now commercial). We have also had quite a few projects centered on automatic testing. Testing of logic chips, testing signal generators, amplifiers, filters, projects much like an engineer experiences when entering industry.

Bioengineering concepts ---How about characterizing a person with transfer functions?

There are many ways that you might stimulate a person, looking for a specific response. The person could be characterized by how well he responds to the stimulus. Stimulate him with sound (as Marleau recently experienced in hearing testing), then check on his reaction to level of sound at different frequencies. Check on the repeatability of these observations. Stimulate with sound, light, pattern recognition --- and look at speed and or accuracy of response. These are important concepts for pilots, machine operators, etc.

Instrumentation systems

How about developing dedicated instrumentation systems? Marleau has a wide variety of sensors and actuators. How about special purpose signal evaluators? How about checking the frequency response of sensors? The linearity of sensors, the linearization of nonlinear sensors? The above is essentially dealing with individual sensors. Consider many such sensors in some instrumentation system making some composite measure of the system.

Optimization systems

Suppose that we have a system with many set point parameters that can be adjusted resulting in the modification of one or more outputs. How would you strategically adjust these many inputs to drive the outputs to desired values as quickly as possible? Have you ever tried to balance an electronic bridge in AC circuits?

Computer Graphics

A good chance to use your matrix manipulations in moving three-dimensional projects through space with perspective.

I could go on here for many pages citing many project concepts, but this was meant to demonstrate what is a quality project compared to one that is not really of ECE 468 character. Second, we indicated the broad range of topics that might be considered and how narrow and specialized a project might be. It is important to make the computer do a comprehensive bit of computation for you, i.e., not simply act as a switch or an operational amplifier.

I want all of you to come to my office to talk of project interests. It is especially important to establish ties with potential team mates, i.e., if you are going to approach your project as a team. Teams are especially attractive to modern employers. A team, however, must perform as a cooperative group, all contributing significantly and equitably to a common goal. I expect each person to detail just what his contribution is to a team project. The entire team will receive the same project grade unless I receive private notes from people indicating a specific deficiency in the contributions of a "team-mate". It is important that I be made aware of this situation as soon as possible. In this manner I may be able to help the team come together as a productive unit --- or if necessary, better recognize the lack of performance so that this can be taken into account when considering grades. Any such notice (letter) will be held in the strictest confidence. I want to spend no more time on these potential negative concepts.

Please stop in as soon as possible with specific projects that you might have in mind, or project area thought that I may have stimulated. It is essential to be developing projects well before the introductory lab exercises are completed. I would like to think that you have your project well in mind by the end of the third week of class, and in near final plan form to be presented in your formal proposal due near the end of week six. This proposal should include schematic presentations of all parts of your project, some means of presenting program flow, e.g., a flow chart, responsibility of potential team members, a production schedule, etc. This should be in a form that can be used to determine project progress. No matter how much we stress the importance of digging intently into the project in the middle of the course, we find the computers quite idle during weeks eight through approximately twelve, and then all h--- breaks loose rushing at the end, doing a less than quality piece of work.

When it is realized in just what direction you people choose to direct your project, the format of the syllabus may change extensively. Please do not think of the syllabus as a contract, but as a dynamic schedule subject to modification.

All Comments and Questions to

Shoaib Junaid at junaid@cae.wisc.edu