This semester I’m taking 6.131, a power electronics laboratory at MIT. It’s awesome.
Our second lab was building the driver circuitry for a 30A motor. You basically drive the motor with a square PWM wave that is smoothed out by an LC filter. The PWM wave is huge, though, 36V with the potential to draw 30A, so you need to be careful that you pick MOSFETs that can dissipate that kind of power. Similarly the inductor and capacitor are HUGE.Giant inductor and capacitor. We do all the calculations to select the core type and number of turns, and then lucky us get to wind the inductor ourselves. 🙂
We end up using this board a lot. It’s power MOSFETs in a “totem” configuration, where you have two MOSFETs in series, the high side connected to power and the low side connected to ground. By switching which one is on you output from the center of the totem either power or ground. Square wave! You can hardly see the MOSFETs in this picture because they are wearing giant heatsinks. These guys get warm when you try to push 30A through them.
Drive circuitry for the go-cart. We use inverters and comparators to make our own variable duty cycle PWM signals to send to the MOSFET drivers.
Our second big project was making a boost and a buck converter to amplify an audio signal. It’s meant to run off of 4 AA batteries, i.e. 6V. The boost boosts the voltage up to a steady 15V to power all of the power circuitry and then the buck bucks it down to a variable 0 -12V as a volume control for the speakers.
Again, did all the calculations for designing the inductor and choosing the capacitance. You can see that this board has the buck and boost on the same card. The control circuitry included the pot that was able to control the volume.
One of the problems I had with my amplifier was that I couldn’t get the volume all the way to zero. Even when I set the input voltage to zero there was enough noise in my signal to hear the audio coming through (albeit poorly). You can see here an image of a scope of the “zero volts” I was putting across my speaker. I think this was mostly an issue with noise in my boost converter, but I didn’t really know how to make it better. I’m pretty sure my LC filter was designed to spec, so it might had to do with noise inherent in my control circuitry (on a crappy breadboard) or some other ambient noise that could only be fixed with active feedback.
Our most recent lab was all about light. We designed an AC light dimmer, a lamp ballast to run a fluorescent lamp, and a giant booster to run a 200V flash strobe.
My lamp ballast. This guy is actually pretty crazy. The fluorescent lamps that are everywhere require a huge voltage, maybe 400V, to strike and turn on, but once they’re on you want the voltage to be much lower. The lamp ballast circuit is basically an LC circuit that you drive at resonance — when the lamp isn’t on yet there is a huge resistance ( the lamp is in parallel with the cap ) so it runs at high gain as if the lamp where an open ( a gain of around 16 for us ) and drives up the voltage. Once it strikes, it’s has a finite resistance and damps your LC circuit, pulling the gain down to maybe 2, which gives it a reasonable steady state voltage. Awesome.
Drive circuitry for the AC light dimmer, the lamp ballast, and the flash strobe.