Yesterday I got to find out I've been an idiot for many years. It's one of those things where I knew everything brilliantly except for one simple fact…
When I was in grad school, my sister was dating a brilliant Oregon State University engineering undergrad named Simon. Simon was headed to the finals of a national engineering competition with a simple goal: build a machine that can lift the most pennies (plus the machine itself) out of a foot-deep box in 60 seconds. The only power source allowed: a single Radio Shack D cell powered by a small Radio Shack DC motor of specified number. I kibitzed Simon to a national win by a factor of two over his nearest competitor.
Simon asked me for advice; I saw no ethical reason why I couldn't give it. The trick is simple. Energetically, you want the battery to be completely drained in 60 seconds, with all of that energy dumped into the motor. Unfortunately, a 1.3V battery discharging into a small DC motor like this won't work it hard enough to come even close. The solution? Boost the motor voltage! Build a boost converter that takes the battery voltage up to 15V or more, and spin the motor like crazy.
Simon went off and did some tests with a motor and DC power supply lifting a weight. He found that about 15V was optimal, but opted for about 5V to reduce the likelihood of burning up the motor.
Then it was time to figure out how to build a boost converter for a single 1.5V battery. Remember, this was many years ago: these converters weren't commercially available.
The key component in a boost converter is the oscillator: there's no good way to boost DC. The fundamental problem is that a "diode drop" is about 0.7V. This doesn't quite leave two diode drops worth of slack for building something. Thus, things like the classic diode-capacitor ladder don't work. Even a transistor doesn't work, because there's a diode drop between emitter and base and between base and collector. So I designed a relay and transformer circuit (!), then promptly broke a lead off the only transformer Simon and I had. So I sent him on to the Electrical Engineering Dept. at OSU, and wished him luck. The ECE guy designed some circuit, and I heard no more about it until reading in the paper that Simon had won. (I received no credit, which was disappointing.)
So I'm staring at a diode-capacitor ladder voltage tripler in a circuit one of the PSAS folks designed yesterday, and I'm like "hey, how do you deal with the diode drops if this voltage were lower". Everyone just stares at me, and then they all blurt out in unison: "just use a FET". Just use a FET. A Field-Effect Transistor works in such a way that there's no diode drops involved. I could have done it back then, if I knew what I was doing.
One simple fact.