In setting up the actuator, I realized I've never known the difference between unipolar and bipolar stepper motors, and that I've never used a bipolar one before. Steppers don't spin like regular motors. They have several repeating magnetic stops inside, and you run current through separate wires in sequence to make them "step" to the next magnet. They all have an angular rating - a degree value they'll spin per step - and mine are 1.8°, a very typical value that yields 200 steps per revolution (360 / 1.8). The unipolar has 4 circuits - 4 input wires that run to usually one or two ground wires through separate windings (coils of wire that act as electromagnets). When you apply power to each sequentially (looping through them), the motor continually jumps 1 step in that direction - reverse the order you pulse the inputs to step the opposite way. In a bipolar, as I have learned, you need 8 transistors (electrically powered switches) as they're are only 2 circuits through the 4 wires (1 through one pair, 1 through the other), and no grounding wires as in unipolars. Basically you're turning on 2 transistors to open a path to source and ground through a pair of wires, but at some point, you're opening a pair that lets current through the same pair in reverse - hence "bipolar." To do this you need a transistor from each wire to both supply and ground. I'm not sure why anyone thought this was a clever idea, but at least I finally made sense of the datasheet and got it working! The 8 transistors I used are from a little bulk purchase I made a long while back, and I'm glad I did. They're NPN Darlington transistors of the TIP120 variety. Darlingtons are actually packages of 2 transistors in a setup that blocks current spike backlashes from the motor windings which could damage the microcontroller. Speaking of, I'm using a BASIC Stamp BS2p40 here because it's so fast and easy to prototype with, but at $100 ($90/ea these days), I'm keen not to damage it (any more - I've blown a few of it's 32 input pins before). The only 3 commands necessary with the stamp, aside from GOTO for the loop, were HIGH, LOW, and PAUSE to power the TIP120s, and subsequently, the stepper windings.

The surface of the dish is actively controlled by a few thousand linear actuators. The original design used a feedback loop involving laser ranging on the ground, but that didn't really work out. Now they monitor temperature and position and compute the correct actuator positions, with some feedback from peaking up the signal.