Motorized rotary switch

Hi - I'm pretty much new to motors, so I'd like some advice about the best way to approach this problem.

I have an 8-position rotary switch that is, most of the time, manually controlled. But, in certain circumstances, I'd need its position to be modified based on software input. So the first thing that I thought was to replace the switch's shaft with another one that would be connected to a servo.

But then the question is, does a servo motor allow for manual control when it is not operating? Won't it just lock in a particular position, preventing the user from manually using the switch? Is there a particular type of servo that I should look for? Or maybe this approach is wrong to begin with? :slight_smile:

Thanks a lot!

wotaskd:
But then the question is, does a servo motor allow for manual control when it is not operating?

For typical R/C servos, the answer would be "no", as long as power and position signal is applied; if the power is removed (and the positioning signal - you should never apply a positioning signal to a servo without it being powered; it can damage the servo - some servos have protection against this, but not all). If there is no power supplied to the servo, the servo can be "back-driven", but this is bad for the internals of most servos (stripped gears being the most common).

wotaskd:
Won't it just lock in a particular position, preventing the user from manually using the switch?

Only if power and signal is applied - see above.

wotaskd:
Is there a particular type of servo that I should look for? Or maybe this approach is wrong to begin with? :slight_smile:

I would think if such a device existed, it likely wouldn't be a standard R/C servo (or use the same signals), and it would likely be very expensive.

Back in the day, when televisions had knobs and remote controls came on the scene, there were motor controlled channel changers; they allowed you to change the channels on the TV, but you could still manually switch the channels if needed. I am not sure how such devices worked, but I imagine that they had some kind of integrated electro-mechanical (or similar) clutch that would engage the switch when turned on, then disengage (to allow manual control) when they turned off. In effect, probably something similar to how the starter motor on a car works (the gear on the starter motor only engages the flywheel while starting the car; there is a solenoid that controls the position of the pinion on the motor - though some motors use a completely mechanical method as well).

So - perhaps look into how these systems do it; I would have to say, though, that this would be something you would have to build yourself, and it wouldn't be easy nor inexpensive. There may still exist motorized multi-pole switches; but again, they are not likely to be inexpensive (if they can be had at all).

I am curious why you are going for such an electro-mechanical method in the first place, and not simply using a digital solution of some sort...?

Is there a particular type of servo that I should look for? Or maybe this approach is wrong to begin with? smiley

Why not replace the rotary switch with something like an encoder connected to your arduino. The operator turning the encoder tells the arduino what to do and software could also tell it what to do. You may need some form of feedback to the operator so they know what position the switch is in when they turn it or software alters it.
Something like this Rotary Encoder - COM-09117 - SparkFun Electronics with feedback like this Rotary Encoder LED Ring Breakout Board - Blue - COM-10407 - SparkFun Electronics

It's possible to get motorised fader controls that present essentially the same problem, so I'm sure there is a solution.

How determined are you to retain a physical rotary switch? It ought to be possible to design a ratchet system that enables you to nudge the switch left or right one position at a time while still allowing manual control. But it might be tricky to design and build.

Perhaps a better approach would be to separate the manual input from the position display. I mean have a position indicator which is controlled electronically (digital display, or physical dial/gauge moved by a stepper motor/servo etc) and then have a separate rotary encoder/switch which the user can move to trigger the controller to change the switch position. So the user can trigger a change by spinning the dial, but the controller can also do it spontaneously, and both of these use the same mechanism.

PeterH:
How determined are you to retain a physical rotary switch?

Very much so. I need to basically simulate a device that exists in real life.

I'm building a 737NG cockpit for flight simulation with panels that look exactly like the real airplane's. The control that I'm trying to emulate is the autobrake switch. It has multiple positions that you select manually while on the ground (brakes 1, 2, max etc), but once airborne, the switch resets automatically to off. It has to be a rotary switch because that's how the actual thing works.

wotaskd:

PeterH:
How determined are you to retain a physical rotary switch?

Very much so. I need to basically simulate a device that exists in real life.

I'm building a 737NG cockpit for flight simulation with panels that look exactly like the real airplane's. The control that I'm trying to emulate is the autobrake switch. It has multiple positions that you select manually while on the ground (brakes 1, 2, max etc), but once airborne, the switch resets automatically to off. It has to be a rotary switch because that's how the actual thing works.

How about a servo armature under the switch shaft (but not attached) that when rotated fully one way pushes the switch to the off position and then is rotated fully the other way to prevent it from being fouled when the switch is turned by hand

Untitled.jpg

Riva:
How about a servo armature under the switch shaft (but not attached) that when rotated fully one way pushes the switch to the off position and then is rotated fully the other way to prevent it from being fouled when the switch is turned by hand

That's a great idea! I'll try this.

Thanks a lot!!

A servo can control a rotary switch like below. If the switch has "click" positions, they may pose a problem for a small servo. You might try a small metal gear servo to see if it can move the switch. In my experience with servos removing the control signal to a powered does not normally cause the servo to do wierd things. The effect might be tested by attaching and detaching a servo for several seconds each in a loop to see the servo reaction.

http://web.comporium.net/~shb/switch.htm