Aerial rotator needs position sensing.

I have cheap, NOS aerial rotator with a broken controller. The way it works is pretty primitive... the motor runs 24V DC with polarity determining the direction of rotation. The controller has a motor geared down to run at the same speed to give crude synchronisation. Due I assume to its age, one of the plastic controller gears has shattered so I intend to make a new controller with a Nano. To do that, I need to sense exactly where the rotor is and ideally add end limit switching besides the software limits.

Under the rotator there is the rotor spindle with not much exposed and the motor with not much access to its spindle.

What ideas do you have for doing the sensing please?

Most old-school rotators have a switch that triggers every few degrees. The user turns the knob on the controller and the controller counts the pulses on the switch wire until the antenna has caught up with the knob position. Rotation of both the control knob and the rotator is limited to 360° so turning the knob to the limits in both directions will re-synchronize the antenna with the counter mechanism.

It shouldn't be hard to figure out how the feedback signal is generated in your device.

There IS no feedback, that's the point. It uses a motor in the controller running at the same speed, using gearing.

Do you see the brushes on your 24VDC motor? Neither do I. I have exactly the same rotator and controller, only so much older. The motor is a split-phase 24 VAC motor. The series capacitor is switched by the controller to control direction. That is why there are only 3 wires going to the rotator!

Paul

Ah, thank you Paul that is a great help. Looks like I perhaps should be looking to mod the controller rather than the motor housing then. I will have to give it some thought.

These rotators used to be dirt cheap, but secondhand ones fetch silly money now, so its worthwhile to mod it.

Remove the external donut-shaped gasket, and the internal bolts holding the motor plate down, and you can remove the entire motor/reduction gear assembly. A 10-turn pot with a properly sized gear to mesh with the reduction gear just prior to the final output gear makes a good position indicator - think I found a gear from an old inkjet printer years ago that meshed nicely.

Alternately, you can just try to simulate the original indicator by timing the length of time power is applied to the motor. Been years since I actually used an antenna rotor, but if I recall correctly you could calibrate the indicator by running the rotor till it reached the hard stops at each end of travel. The motor is not particularly high torque so it can withstand being stalled for short periods of time.

Never got around to working on it, but had contemplated years ago intercepting the IR signal between a Tivo and a digital converter box to automatically rotate an antenna to the correct position when the channel changed.

Thanks for the suggestion David, it co-incides with some of my thinking. I think that mine has no hard stops either, so will need to incorporate some. Otherwise I could just look out of the window maybe :wink:

If the original controller used timing, arguably you could use the same with the Arduino.

Put a magnet and reed switch on the (large) output shaft for zero position..

Paul__B:
If the original controller used timing, arguably you could use the same with the Arduino.

Put a magnet and reed switch on the (large) output shaft for zero position…

But then you would need more than the three wires currently used by such rotators. With mine, I have to run the controller to a stop and then go look to see if the rotator looks like it is at a stop and then back up the controller to the position I want. Really cold weather makes the rotator move slowly, so winter, it it always off sync.

Paul

I bought a secondhand Hirschmann controller, believing that they are universal. They are not! The Hirschmann does a complete revolution in about 60 seconds, my rotor needs 77 seconds.

I have taken the motor apart as suggested and have ordered a 10 turn pot and some Hall effect sensors to experiment with. However I am minded to make an electronic version of the mechanical controller now to save the hardware modifications.

It would appear that if I run a timer for 77 seconds, I will be able to synchronise the two together. My idea is to utilise a Mega with display shield I bought months ago and never used. A nice graphic compass display with a pot to set the heading required. Apart from the graphics, it shouldn't tax me too much hopefully.

I have now ordered a three axix compass module to play with. One way or another, this is going back into service. :slight_smile:

You will have to mount it on an arm to substantially distance it from the mast!


A plastic arm of course. :grinning:

Paul__B:
You will have to mount it on an arm to substantially distance it from the mast!

And keep it away from the magnets, electromagnets, and steel parts in the motor.