|I watched lecture #1.
I read chapter 1 of Foundations of analog and Digital Electronic Circuits.
Professor Anant Agarwal’s first words in the first lecture of this class are, “So, one question to ask ourselves is, what is engineering?” I took that to be a good sign, because that’s a question I was really hoping the class would address. “Engineering is the purposeful use of science,” he quoted from the textbook. It’s a satisfying definition that puts that question to rest for me. It rings true.
The lecture follows chapter 1 of the book pretty closely, though the emphasis wasn’t on all the same points. The lecture also spent some time on a topic which, in the book, was relegated to an appendix. And also, the lecture included some very cool demonstrations!
In both the book and the lecture, the concept of levels of abstraction was central. This is a concept I’m very familiar with, but one of the presented levels of abstraction took me by surprise: that scientific laws are an abstraction of experimental data. And, even more insightfully, an abstraction between scientific data and engineers. The point was made that you could successfully engineer some systems without scientific laws by relying on a direct measurement for every prediction you have to make. But that is very limited and, for all but the simplest machines, prohibitively difficult. So we create the abstraction of scientific laws not to complicate things, but to simplify them.
The abstractions continue, layer upon layer, until you’ve got things like desktop computers running operating systems and complex software. This particular lecture (and chapter) focussed on a very specific abstraction. It’s the abstraction that allows us to make the leap between Maxwell’s equations and the far simpler set of rules that are used to analyze electronic circuitry. That was the the next good sign. That’s the really big thing I wanted to grok.
The abstraction in question is the “lumped circuit abstraction”. This abstraction is made possible by the “discretization discipline” or “lumped matter discipline”. It’s actually something so simple and fundamental that it’s easy to take for granted. We all like to “lump” the world around us into discreet objects. Pencils, paper, chairs, books, air, etc…. And we expect these objects to interact with each other only in certain ways – through certain interfaces (e.g. pencil writes on paper). But they aren’t really as separate as we treat them. They’re all governed by the same laws of physics applying continuously over space and time. We create conceptual boundaries between them that are true enough for our purposes, but that don’t hold true under all conditions.
We discipline ourselves to limit our implementations to machine components for which certain assumptions hold true … so that our neat and predictable boundaries and interfaces will hold true. But an engineer needs to understand what those assumptions are so that he or she knows when the abstraction does not apply.
It occurs to me that this business with boundaries rubs up against some big philosophical issues. Another boundary of that type might be the boundary between self and other.
The class and the book go into some detail about this abstraction and the various tools that emerge from it. If you’re really interested, you should take a look for yourself and consider doing this along with me!
Preparation for Labs
I’ve ordered or made arrangements to borrow most of what I’ll need to complete the labs. Some of what I ordered has arrived, other things I’m still waiting for. As far as the more expensive pieces of lab equipment go, I’ve ordered a very fancy multimeter that is also an oscilloscope. It has a small LCD to display its plots and it can also send the data live to a computer via USB. It should cover most of my oscilloscope needs, but I believe I will at least once require features it doesn’t have. For those, I should be able to borrow the use of a very feature-ful oscilloscope for long enough to do what I need to do.
For the function generator, I plan to use the audio output of my laptop and an adjustable amplifier, calibrated using the multimeter (although the calibration may turn out to be non-linear). I’ll write a simple program to generate the signals I need.
The final lab also calls for a ready-made circuit board that was provided to the students at MIT. It consists of a 16-bit counter chip and a 16-bit-address 8-bit-data memory chip. In my big Mouser order, I included those things so I could replicate that. The tricky part will be programming the memory chip (CMOS). It comes with its data already on it at MIT (probably an amusing audio message?). I plan to borrow the use of a multifunction DAQ and to write a program that uses the DAQ’s digital I/O lines to program the chip.
The multifunction DAQ can also be used as a function generator (within -10v to 10v) if the audio output trick becomes a pain. The multifunction DAQ’s analog output would also be linear and automatically calibrated. …but I probably wouldn’t be able to use it at home.
Thank you for reading!
Course citation: Agarwal, Anant, 6.002 Circuits and Electronics, Spring 2007. (Massachusetts Institute of Technology: MIT OpenCourseWare), http://ocw.mit.edu (Accessed 23 Aug, 2010). License: Creative Commons BY-NC-SA
Book citation: Agarwal, Anant, and Jeffrey H. Lang. Foundations of Analog and Digital Electronic Circuits. San Mateo, CA: Morgan Kaufmann Publishers, Elsevier, July 2005. ISBN: 9781558607354.