In order to test my hypothesis regarding the use of PWM controlled electromagnets arrayed within a grid structure I have made a prototype. The Arduino Duemilanove board that I am using has 6 native Pulse Width Modulation outputs, so that is how many I used. However a better test would have been with 7 coils – one in the middle, surrounded by six closely placed coils – which would create the maximum number of coils for a given area. This would better confirm the validity of the experiment, but I believe that 1 + 5 is reasonably useful.
The coils are coupled to the low current 5V Arduino outputs via a high speed amplifier with a MOSFET output stage that was modeled in LTSpice and hand soldered point-to-point.
I initially powered the prototype with a 180W SMPS from Hypex, but it was current limiting – only allowing the the device to work in short bursts. But work it did! Consequently, I quickly hooked up a linear power supply, and keen to proceed at pace placed the only fuse I had to hand – 10 A. It again worked for one glorious second – with excellent magnetic flux – before not working. It happened a day before a friend arrived from overseas for our Christmas/summer holiday, so I haven’t as yet had time to unbug it as I have just returned, but I am assuming that I have cooked a coil, and hopefully not the MOSFET outputs.