Thursday the students turned in their final design reports (which I have not read), and showed me their prototypes.
None of the groups had working prototypes, because all of them started too late on acquiring parts and building. I will have to create and enforce a schedule next year that gets them started on prototyping much earlier. A lot ran into the problem that electronics parts need to be mail ordered, and that introduces a few days delay. Others ran into problems they had not expected (like that styrofoam coolers can’t be sold in Santa Cruz—a problem for the incubator group).
I went with two of the groups over to an electronics lab to help them debug the prototypes that they had started.
The incubator group had a thermistor temperature measurement circuit and Arduino program that was always reporting freezing temperatures. I walked them through measuring the thermistor resistance (seemed ok, and changed if the thermistor was warmed by holding it), measuring the voltage from the voltage divider (seemed ok, and went up if thermistor was warmed), looking at the raw reading from the Arduino (again, looked ok), and looking at the reading-to-temperature conversion routine (did not have a return statement, so was always returning 0).
The PCR machine group had two Peltier devices, one of which seems to have been damaged (open circuit with multimeter). I helped them debug their H-bridge circuit—they had it working when there was no load (though they had not verified that it did so), but the Peltier device was too big a load, and the H-bridge had a short-circuit protection circuit that turned off the H-bridge after 50μsec. The Peltier device needed about 10A, and the H-bridge shut down at 8A. Their power supply was only capable of delivering 1.5A anyway, so the Peltier device would not have cooled or heated the block fast enough anyway.
The centrifuge group had demoed their project earlier in the day. Their design was primarily mechanical, and they were still having troubles with it not being stable. Based on the vibrations of their motor, I think that they were only getting up to about 600 RPM, not the 3000–4000 RPM that they wanted. The limitation may have come from an unbalanced rotor.
I think that next year I want to discourage mechanical design projects, because we don’t have facilities to do much construction and debugging of mechanical designs (and I have less expertise to help them). Electronic design and programming are fields that I can help them more with and which we have better facilities for.
The biggest changes for next year are
- more programming exercises with Arduino (including thermistor or phototransistor)
- start project sooner
- weekly lab time
- deadlines for parts ordering
- maybe enforcing one design project for entire class (with different teams trying different approaches)