I just started my first class on Coursera.1It’s Introduction to Power Electronics from the University of Colorado. The tools my team designs and builds help our customers validate their voltage regulator designs. This class will help me understand a little better what our customers go through. When I attempted the first homework assignment, I pulled out the tools I typically use at work: 0.9mm mechanical pencil, blank photocopier paper, and the calculator on my Android phone.
Nope! That just wasn’t going to cut it.
Off to the office supplies store I went and got the real stuff: 0.5mm mechanical pencil, an eraser in a holder, engineering paper, a D-ring binder, and enough dividers. When I got home, I retrieved my HP calculator.
And although those tools got me through college, it wasn’t going to be enough now. I had to get my reading glasses, too.
Now I’m ready.2And I’m having fun. And I hate to admit it, but I’m working hard. Where are my college buddies so I can relax with a game of hearts? ;–) :-)
Sunday was a big day! Jamison and I started with another trip to Fry’s. We needed more magnet wire, a better power supply to clean up the noise we saw on the cheap one we already had, and some perfboard. On the way home, we stopped by Home Depot for 6″ pipe and Fred Meyer’s for reading glasses. Never needed those before… B-)
Upon our return home, we set up our lab and pulled out the new power supply. After deciding to use a different scope probe and calibrating it (we missed that step last time), we got a nice, clean 5V voltage. Because we wanted to see the contrast, we then connected the probe to the cheap “wall wart” power supply. Look at that — a nice, clean 5V voltage! Huh. I guess the new power supply can go back and I’ll try to be more understanding when a customer calls for tool assistance and hasn’t calibrated his scope probes.
Jamison wrapped a second coil and we hooked it up to a green LED. Connect the power, move the coils close together. First light! Yes!!
Now it was time to improve Jamison’s project.1 We tried a full-wave rectifier to increase the LED’s brightness, with not a large (if any) effect. We built a two-stage Cockcroft-Walton generator, with much better success.
Next we decided to improve the coils. We cut the 6″ pipe into short lengths, an inch and a half or so. We wrapped the magnet wire around them and used the oscilloscope to tune the number of wraps to get the desired frequency. It seemed like we were having to remove too many coils, so we found a formula for the inductance of a short air-core cylindrical coil. Plotting inductance as a function of wraps showed we could get by with much fewer than the 18 wraps we thought we needed for the desired 53 μH.
Emboldened with new information, we continued experimenting and found twelve wraps got us nearest to the desired 80-kHz frequency. We built a second, matching coil and we found we could transfer power much farther. The first set of coils had to be nearly touching. The second set could be more than four inches apart.
Unfortunately, we hadn’t left long enough leads on the coils, so we decided to wrap another set using all the information gained along the way. The forms were only an inch “long.” We sawed deeper groves for the wire’s start and end so the coils wouldn’t need to spread and we spaced tape around the coil to hold them tight. I haven’t measured, but I think we can have six inches between the coils and still light the LED.
We want to try making a smaller secondary coil, maybe an inch in diameter. That will be for another day.
A couple of weeks ago, we last documented Jamison’s progress on his engineering project.1 Much has happened since then.
Decide on prototype
We thought we’d decided what project to build (I thought we were building Wireless Electricity), but Jamison changed his mind. After further research, he decided to build the Low-Power Wireless Charging he’d found on Instructables.com. I’m pleased. We’ll have more control over this project.
Buy materials for prototype
Once we knew what project to build, we thought it would be easy to buy the parts. Wrong!
Buying capacitors is a confusing process. The units are insane. The instructable specified a 100 nF capacitor. That sounds easy enough. Unfortunately, Fry’s shelf labels would list it as a 100000 pF or a .10 MFD. That’s if they actually had it. After trying to find the right caps, we decided to do our shopping online instead and headed over to digikey.com. We had all the parts we needed the next day. :-)
Prototype building went well, at least for the circuit. Once Jamison understood how the little holes in the breadboard are connected, the oscillator went together fairly quickly.
The first coil went together fast, too. The instructable said to do 18 wraps of 15 cm loops. Jamison wrapped the coil around a form, about 4.5 inches square, and then rounded the coil by hand.
Tuning the coil was a problem. The oscillator was designed for 80 kHz. We should have been able to tune the coil by connecting it to the oscillator and removing coils until the oscilloscope showed something near 80 kHz. Famous last words. Instead of a nice sinusoidal signal, we got a noisy, stepped signal. Removing coils didn’t seem to do anything. Maybe the coils were separating, so we wrapped it in electrical tape. That made no noticeable change. Maybe the cheap power supply was causing the noise. Fry’s could rescue us with a nice benchtop power supply.
Time to call it quits and figure out our next step.
It’s hard to beat cheap, so we’re going to attempt the Wireless Electricity (for under $20) project, though that cost estimate seems to assume already having a stocked experimenter’s lab. This is sending us down the bifilar winding rat hole and into Tesla territory. A lot of strange stuff (and denizens) there.