Second lab was smoother than the first

As predicted last week, today’s microphone lab  was shorter than last week’s thermistor lab—almost everyone finished within the 3-hour window.

I do have to make some changes in the lab handout for next year:

• The handout should discuss how to set the current limits for the power supplies.  My co-instructor insisted on our going around to each bench and helping students figure out how to do that.  He’s seen too many blown fuses on ammeters in student labs.  In fact, one of the benches had an ammeter with a blown fuse, because the EE TAs are not trained to make everyone set current limits the first time they use the power supplies, and someone inevitably wires up an ammeter like a voltmeter.  Since the bench supplies default to a 6A current limit, and the ammeters are fused at 3A, a use is blown.  It is not unusual for both ammeters on a bench to have blown fuses, as students switch ammeters rather than fixing their circuits.  I’m pleased to say that we did not blow any ammeter fuses today.
  I should also include some information about how to hook up wires to the binding posts on the power supplies.  We don’t have banana plugs, so students need to strip an inch from the end of the wires and thread the wires through the vertical holes in the binding posts, then screw the cap in place to hold the wires.  One group had just wrapped the wire, and not stripped it far enough, so the binding post was just squeezing the insulation and not making contact with the wire.
• We did not do the DC-blocking capacitor at the end of the lab, because we have not gotten to the theory of capacitors yet.  I think we’ll probably cover sinusoids and complex impedance on Monday.  I’ll cut that from the lab, but add it to some later lab (probably the op amp lab, where it becomes important).
• There is no point to trying to measure the amplitude at the microphone when playing signals from the signal generator through speakers.  With 5 or 6 loud loudspeakers all going at once, the noise levels in the room are too high for any sort of careful measurement.  It was more a time for students to play with the oscilloscopes and the function generators.  I need to come up with a more playful task for this part of the lab—perhaps sweeping the frequency up and noticing at what frequencies the microphone amplitude is largest.  There is a pretty obvious peak at the speaker resonant frequency (around 200 Hz), and the amplitude goes up at the high frequencies.
• I need to explain how much of the wire to strip for the breadboards—some of the students had stripped over an inch of wire, which could easily result in shorting adjacent wires together, but 5mm is a more reasonable amount to strip.

As I planned last week, I gave each group of students the option of turning in the lab report either on Monday or on Wednesday, with no difference in credit.  Everyone chose to have the extra time for the report, rather than early feedback.  Maybe next week some will choose the Monday option instead.

I’m very pleased with the 78¢ speakers (well, $1.28 with tax and shipping) , which can handle 10W RMS (20W max).  With 8Ω speakers, 10W is about 9V RMS or 12.6V peak-to-peak, so no one was likely to fry their loud speaker hooking it up to the function generator.  I understand that the usual tiny speakers that are provided in the EE courses are only 0.25W speakers and are routinely destroyed in the circuits course—and they’re not significantly cheaper.

I think that all the students now know the basics of using a breadboard, though most of them are using wires that are too long, which will become a problem for them when we move to more complex circuits.  This lab called for a trimpot, a resistor, a microphone and the Arduino for getting the current vs. voltage plot for the  microphone.  I think that everyone got that circuit wired up and working and gathered good data with it (hundreds of points worth of good data, with currents from 1µA to 200µA). I think that the students got a good appreciation for the value of the Arduinos for collecting data points, as most ended up with several hundred data points for a minute or two of collecting data, rather than just the 10 points they got with hand-collected data. Some of the students were quite excited to be able to wire up the circuit, collect data, design an appropriate pull-up resistor, and see sound on the oscilloscope.  I was pretty excited myself at how much they were learning and getting done, even though I’ve done the lab myself several times trying to tweak it into something they could do with really minimal prior knowledge.

I worked individually with a few students to help them get their gnuplot scripts working—the lectures I’d given in class had not been very successful in getting them to learn gnuplot, but about 10 minutes of one-on-one (or one-on-two) tutoring got them to the point where they understood what all the commands they were using did.  I suspect that they could have learned on their own, but lacked the confidence to do so.  I’ll be doing more individual tutoring with gnuplot after class on Monday, but I think that several of the students now have enough understanding to help out others in the class. Some of them even got the point of providing reasonable initial values for the parameter fitting, as a working script broke when they rescaled from amps to microamps.  Adding a reasonable initial value (which was 6 orders of magnitude different from the previous fit) made the fitting converge again.

I hope that the students remember to include all the data on their plots, including the hand-collected data at higher voltages.  I also hope that they can figure out how to get the parameters they fit to appear in the key for the plot (the gnuplot script for the first lab showed them how to do it, but some are having trouble abstracting from that example to create a new example).

The maintenance of the multimeter leads and scope probes in the lab left a lot to be desired.  We had at least 2 bad sets of multimeter leads (broken wires at the strain relief of the clip) and at least 3 bad scope probes (broken ground leads that looked intact). Most of the function generators didn’t have any leads at all, though the BNC-to-clip-lead cables are under $6 each.

On Monday, I’ll do continuity checks on all the scope probes for the analog scopes and try to get the lab support people to repair or replace the broken ones.  Since a pair of oscilloscope probes (of better quality than the ones that the lab supplied) costs only $13, it may be worth adding to the parts kit, so that students don’t have the aggravation of having to deal with probes that someone else broke.  That will only work as long as the lab has scopes with BNC-connector inputs, though.  If they surplus the old analog scopes and use just the digital Tektronix scopes, much more expensive proprietary leads would be needed.  I doubt that as many of those are broken, since it takes forever to learn to use the baroque menu interface on the Tektronix scopes, so almost everyone prefers the end-of-lifespan analog Kikusui scopes.  Certainly the circuits lab, which is the first lab in which students encounter oscilloscopes, would be much better off with a dozen working analog scopes, with the digital scopes moved to senior design labs and other labs where their greater capability would justify the much greater amount of time it takes to learn to use them.

I think I’ve come up with a nice “do now” question for tomorrow—one that they can solve with their current methods, but which becomes somewhat easier after they’ve had Thévenin equivalents, which is the main topic of tomorrow’s lecture by my co-instructor.

In Monday’s lecture, I’m going to cover sinusoids and complex impedance, so that we can do DC-blocking capacitors, RC filters, and the models that we’ll need for the electrode lab.