Let's Make Robots!

Normally locked rotation? Power-off brake?

Thinking about the mechanicals of this a little more, I've been considering servos and steppers.  I don't need speed or precision, really.  But I do need to be able to move in either direction, which I now understand to be complicated with servos.  So I considered steppers, which at least can move in either direction. 

A particular challenge of this application, though, is once the rotation has happened I'd like it to be locked in place to resist rotational forces from wind and chop, and I'd rather that not draw power, making even a stepper with holding power difficult.  So I'm thinking that I need some type of normally locked mechanism.  Something like a ratchet, but I'd still like to be able to move either direction, without full rotation.

Perhaps something like a stepper motor and a pin that can be dropped into the rotation mechanism and retracted using a solenoid? A little googling suggests that I might be looking for a "power-off brake" but these all look very expensive.

Any ideas? Honestly the holding tension doesn't have to be absolute, just enough that the motor can be turned off when in position and have some expectation of holding in non-extreme conditions. Perhaps the initial pressure can always be in place and just have the stepper push through it to move the thing?  That does seem like it would draw extra power, though.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
this solution won't lock the motor solid, but it resists movement pretty hard. Some how short the motor power together ( that is the 2 terminals at the motor, not the actual power used to drive the motor.) I bet it will be nigh on impossible to move a gear motor very far and hard to move a servo. Obvious it will move slowly as the back emf dissipates. I've seen this design used in tape libraries to drastically slow the drop of a heavy tape loader arm. Not sure if this is enough for your project.

Thanks ignoblegnome, that's a cool mod.  I think, though, that back-rotation would still occur and one would be constantly running the servo to re-position it.  Worth thinking about though!

Thanks 6677. Interesting suggestion, a quick wikipediaing suggests that a worm gear drive is "self locking" under some conditions to do with the  lead angle and friction angle.  I can't quite picture that in this application, but I'm thinking one would have a worm/screw setup on the pole that holds the antenna and direct drive that from the servo.  I think, though, that this would only allow rotation in one direction? Or does it allow movement in both directions, but only initiated by the stepper, not by the pole.

Seems also that worm gearing doesn't neccessarily self-lock under conditions of vibration, but this isn't a safety situation so that might be ok.

I need to draw some pictures, I know!

6677's picture

if you fixed the spur gear to the pole (basically had teeth going around the pole) the worm gear would be mounted to the unit that is moving around the pole. this would then work in both directions and be self locking. The amount of force required to get a worm gear assembly to move either has to be a pretty high frequency vibration or a force strong enough to break the teeth usually. Most of the time the correct circumstances are only achieved when your trying to break it.

Worm drive is sounding like the way to go :) Mechanically I'm considering a two-part pole, where the bottom half will be fixed and the top half will rotate.  The antenna and micro-processor are in the same box, so I was planning to attach that permanently to the top half of the pole and have a small motor at the bottom of that box, linked to the fixed portion of the pole, so it's rotation will move the top half of the pole (moving the motor around the fixed point).  Just as you described, 6677.

I've spec'ed out a nylon worm gear and matching worm from SDP/SI.  Man, they have a lot of stuff. e.g: Worm Gear: A 1M 6-N24040 Worm: A 1M 5-N24.

Not entirely sure how that gets linked to the stepper motor, but going to combine that with the Pololu stepper motors and the Driver carrier from Pololu. Is that overkill?  I can take the direct 12VDC for motor control rather mess around with feeds from the board.

Hmmm, I probably need to ensure that the stepper motor shaft is the same size as the bore on the worm part.  5mm.  Hmmm, but I don't know the length of shaft that I need.  Presumably I can find some type of shaft extender?

 

I've spec'ed out a nylon worm gear...

What made you pick nylon?  I'm thinking you'll want to reconsider your material selection.

Maybe I'm just afraid of metal :)  This is going to be outside, so I want something that doesn't need to be lubricated etc.  I'd been checking out bearings and now you mention it I remember that Nylon expands with water.  What would really be great is a worm gear made from the same stuff that the igus bearings are made from. Or maybe not, perhaps I just need to accept that it'll need to be relubricated ever now and again.

Back in the old days (100+ years), folks could design for most any kind of mechanical motion you'd want - without the use of fancy motor controllers, micro-processors and the like.  Could you describe your application requirements a bit more: resolution, speed, repeatability?  It sounds like you're positioning an antenna - is that right?  That's a typical application for a worm drive (which can rotate in both directions).

 

EDIT: I just read through some of your previous blogs and better see your application.  You want the pole stationary and have the antenna element/assembly manuvere itself around the pole?

6677's picture

If your chassis allows it you may be able to construct a worm gear transmission. A worm gear system will only turn when the input (stepper motor would be fine) turns. the output is locked all the time. There is however a large gear ratio reduction with this set-up but this may not matter and can also be geared/pulleyed up.

Assuming the chassis allows it this would be the perfect solution not requiring any extra programming or powering extra braking systems

ignoblegnome's picture

What if you use a servo, and try this modification so you can monitor the position? Then you can power off the servo after getting it into position, and just passively monitor the position. If it changes, you can power the servo back on and reposition it.