Gyros...I've thought about this. I can't offer much advice, but here it is anyway:
I'm building a large model of HMS Dreadnought, and turret rotation feels like a must-have. Using a Basic Stamp microprocessor I can get
turret control the way I want it. (You'll need a flash/shockwave-equipped browser to see the example. Drag the yellow arrow to set the bearing, hit the train button to start the turrets.) Once the system is complete, I'll set a director bearing using one servo, then use another servo to train the turrets to acquire that bearing if they can. They'll stay parked if they can't.
Now, what would be nice is the use of a gyro to fix a director bearing
independent of the heading of the ship, so that the turrets will track a bearing, park if they can't, and acquire it if they can, no matter what direction the hull is pointing, while the turrets are set to train.
I've been told that heli gyros won't work for this - they maybe need more rotational motion than a turning ship has, in order to accurately track (and in a heli's case, control) movement. Whether that's the case or not I don't know...but the price of heading lock gyros is prohibitive for an experiment that might not work.
As a result I'm currently looking into doing this the good old fashioned mechanical way. I've certainly got the room for electromechanical gyros, and I know that would be good enough, in a two axis system - I'm not worried about pitch on a seven foot hull, nor the space for extra batteries & bits - to fix the bearing. Even if it drifts over a few minutes, it would only require sending a new director bearing to reset it now and again.
If someone HAS experimented with hlg's, and can prove they have enough accuracy over small movements and reasonably long time spans, I'll add my request to Peter's as I'd love to hear about the results.
Regards,
Andy