chapter 9, Control Computers and Reconstructors — The Brains

Excerpt

“I think there is a world market for maybe five computers.”

—IBM Chairman Thomas Watson, 1943

Just as the wavefront sensor is the eyes of the adaptive optics system and the deformable mirror is its fingers, the control computer is the brains of the system. While many systems can have just a few wires connecting the output of the sensor to the mirror, some systems have racks and racks of electronics to make the precise calculations necessary to drive the control mirrors with the confusing signals from the wavefront sensor detectors.

Many servo-control systems have been developed over the centuries. From simple pressure regulators to the artificial intelligence of robots, the idea of automatic control of a process was advanced.

While controls engineers spoke in tongues with block diagrams and Laplace transforms and assumed recognition of terms such as gain, type II controller, and phase lag, the adaptive optics scientists were going all-out to coordinate multiple feedback loops, all working in parallel, to drive a multiactuator deformable mirror.

Single-channel servo control

The first step in describing an adaptive optics control system is examining how a beam of light could be stabilized by a single-axis tilt mirror. If we can't control one beam, we can't control a thousand pieces of it individually. One of the nice things about adaptive optics controls is that most of them are handled by simple, or not so simple, single-channel control theory. Then there are just a lot of them tied together. A block diagram of a control channel is shown in Fig. 9.1.