CmeBrew:
Saving money on wire is one thing I had in mind. I have 2 AGW wire to run a 1000 W inverter. It's heavy, expensive, and not very flexible. I didn't think it was very practical to run a 100 feet of this stuff. I also wanted to max power going to batteries. What's the point of making all this effort to build something, if you loose half your power in a transmission line? That's why I opted for higher voltage.
The interface is not as complex as it may seem. I realised early that I would need something to control this puppy. The interface provides flexibility and automation. It's run from the parallel port off my old 486 pc. You can get kits for about $30 which will control 8 relays. All you need to do is print a byte to the parallel port. Each bit of this byte is going to determine if the corresponding relay is on or off. A 0 means it's Off and a 1 means it's On. Google "parallel port relay."
The parallel port has 18 data lines, including ground, and 3 ports (usually ports 378h, 379h and 37Ah). The data output port controls lines 2 to 9 (output only). The control port controls lines 1, 14, 16, and 17 (input/output). The data in port controls lines 10, 11, 12, 13, and 15 (input only). I got two kits and hooked them together so I can control a max of 12 relays.
I run the RPM sensor off of line 10. I made a skirt around the hub of the mill with a notch in it. On one side is an LED and on the other is a light sensitive resistor. I use this to drive a transistor that brings line 10 from high (+5V) to low (0V). Every time the blade makes a revolution, line 10 goes from high to low. I use this as a signal and measure the amount of time between signals to calculate RPM. I can also count the number of signals in a small time interval to get an average. Eventually, I'm going to trash the computer and replace it with something (?). It sucks up over 100 W.
I can post pics if anyone is interested in more details.