More on electronics course design
I wrote earlier this week about an applied electronic circuits course for our bioengineering majors that Steve P. and I will be designing over the summer, and how it will be designed backwards from the labs we plan to have them do. That wasn’t quite an accurate description of our design process though, as the selection of the labs is based on more fundamental curricular goals. I think that Steve and I have similar goals for the course, based on our discussions of it, but we’ve not tried to write them down formally yet. I believe that we’ll need to do this for ABET, if the bioengineering program goes for full engineering accreditation.
I know that the EE Department wants the course to have a wider appeal than just to bioengineering majors taking it because it is (well, will be) required, as fluctuations in the cohort size for engineering majors has been large (biggest in computer science, which has been on a roller-coaster ride nationwide for the past couple of decades). The BME department chair, who is also the bioengineering undergrad director and main adviser, estimates a class size of 30–50 students if the course is offered this coming winter, and expects growth to 50–60 over the next few years. The labs have room for about 20 students at a time, so this means 2 lab sections this year, growing to 3. If we had a wider appeal, we might be able to pick up some computer science students, some game design students, or some Digital Arts and New Media (DANM) students, who are currently avoiding EE courses despite some interest in electronics because of the highly theoretical EE 101 course that is prereq to everything. One problem with this broadening is that most of the potentially larger audience will not have been required to take the physics electricity and magnetism class, which I think is going to be an essential prereq.
Trying to put these two ideas, the need for more explicit goals for choosing labs and a wider potential market for the course, together, I came up with a possible theme for the class to guide lab selection and topics for the course:
Students completing this course will be able to design and build simple biosensors for measuring chemical, electrical, or physical properties of biological systems (especially humans) and recording the measurements for computer analysis or interaction.
I still have to run this idea by Steve, and we’ll almost certainly need to tweak it a bit before we’re happy with it. I’m particularly interested in two words in the goal (which I think Steve will agree with): design and build. I don’t want this class to be a theoretical circuits course, in which students learn to formulate and solve large systems of linear equations, with no connection in their minds to any design problems. I also want the students to learn to make things that they can take home, not just breadboard prototypes, though I may have to compromise a bit on this one, with most labs being breadboard only, and only one or two getting them to the next stage: a soldered prototype. Designing PC boards is probably beyond the scope of the class.
The labs I talked about in the previous post all fit within this theme:
- Skin conductance meter.
- Electrical field measurements in an electrophoresis gel. This one is a bit of a stretch for the new theme, as electrophoresis gels are a standard lab tool, but not a biological system.
- Conductivity of saline solutions. Used for measuring ecosystems (like rivers).
- Do-it-yourself EKG
- Optical pulse detector
- Breathalyzer
- Thermistor-based temperature measurement.
Since that post I’ve come up with a couple of other ideas:
- Capacitance touch sensor (nice intro in Microchip application note AN1101). This one is more ambitious than the EKG even, in that it involves a capacitance-sensitive oscillator design, but the oscillator is low frequency and the frequency sensing can be done by a microprocessor. They could use the same Arduino that they use for recording voltage data in other labs, but we’d have to provide them with a different program, since we are planning on having no programming required in the class or as prereqs to the class. If the oscillator is a low enough frequency (say around 1kHz), then the program could be almost identical to the interrupt-driven data logger that my son wrote for the homemade super pulley, but interpreting the interrupt times differently.
- Potentiometer-based goniometer for measuring joint angles. Like the thermistor lab, this one is mainly a voltage-divider lab. The design problems are more mechanical than electrical, but it would be useful for students in understanding noise problems in sensors, since potentiometers are notorious for the noise in the signals. It would also prepare students well for later work (outside this class) using servos, since most servo motors use a potentiometer-based angle measurements in their feedback loop. I wonder whether the guts of a servo circuit are within the scope of this class—probably not, as much of it is concerned with handling the power demands and voltage spikes of the inductive load of the motor.
There have been a few ideas that I’ve had to reject as being overly ambitious for a first circuits course or too expensive to implement: anything involving a microscope or micromanipulator setup is too expensive (eliminating most of the nanopipette work done in Pourmand’s lab), anything requiring weeks of practice to set up is also out (eliminating much of the nanopore work in Akeson’s lab, since forming the bilayers is still an art that few master quickly, also eliminating most work that would require growing neural cell cultures). The nanopipettes and nanopores also require measuring currents in the picoamp range, which requires very sensitive electronics and extreme attention to noise control (including doing everything in a Faraday cage), which is probably beyond the scope of a first circuits class. A subsequent bioelectronics class could cover patch-clamp amplifiers and other specialized measuring circuitry.
We probably do want some lab that measures ionic current, though, as that is fundamental to both the nanopipette and nanopore work. I wonder whether we can do a scale model of the nanopore (with pin-prick size holes in plastic or Teflon), so that the students can measure the currents without the picoamp scale problems.
I suspect that when we try out the labs, we’ll find that only about half of them work well for us, and that several of them rely on the same basic circuit ideas (like voltage dividers to convert resistance into voltage). In the latter case, we may give students the choice of several different labs to do. I’ve always liked the approach of giving students a number of different design problems (of roughly comparable difficulty and covering the same fundamental concepts) and letting them choose which ones to work on. I’ve never implemented this approach in any of my classes, though, as I’ve either been faced with a pre-designed set of lab exercises that I didn’t have the authority or time to redesign or I’ve gone for quarter-long independent projects which are not even attempting equivalence between projects.
Whether we end up rejecting some of these labs or merging some into alternative choices for a single lab, we’re obviously going to need more lab ideas. Does anyone have any?
Filed under: Uncategorized Tagged: Arduino, bioengineering, circuits, course design, teaching
