Mechanism to move an object in a circle without rotating it

[Infraviolet]

I've been doing some thinking and wondered about your ideas on this. Any suggestions as to how an object could be moved so that its centrepoint, or any other point on it, translated on x and y so as to form a circling motion, whereas the object itself keeps a fixed rotation according to the global reference frame. So we've got the object doing sinusoid motion on the x axis and cosinusoid on the y axis, but not itself rotating. I'm after ideas of how to do this so it can all run from just a single spinning input, so no x-y leadscrew setups with steppers...

I can think of having a bunch of gears arrayed about a centre gear or around an internal gear's rim each with an eccentric crank shaft on them, and all the cranks going in to the moving object, with the crank-attached gears all having the same tooth count then as the ring gear or central gear turned they'd each turn in sync, but part of me thinks this is more complex a mechanism than necessary.

I've also wondered if there was anything one could do from a scotch yoke mechanism, but again think you might need several of them run from a gear chain keeping them in sync so as to preserve the moving object's global rotation.

Is there a really simple trick to this?

[langwadt]

one motor to rotate, two identical pulleys one stationary on the center the other on the object and a belt

[beanflying]

One non/less mechanical option would be to incorporate a simple R/C Heli Gyro stabilizer or it you need a bit more certainty then a heading hold one. Micro, Gyro and some sort of stepper or servo out on the arm. Getting power to it is simple and there is plenty of off the shelf wiper options for not to many $ for the centre.

Buy one for a plaything typically $10-30 for a cheapie Aliexpress Link

Just rereading your first post and if you want to keep it rotating then just get some power out to your object and run a Micro/motor combo to control speed or rotation. Take a read in this topic
https://www.eevblog.com/forum/projects/slip-ring/msg2776882/#msg2776882

[Ian.M]

one motor to rotate, two identical pulleys one stationary on the center the other on the object and a belt

Yes, however the devil is in the details.  A toothed belt (or chain) drive is probably the easiest way of avoiding the object's orientation slowly rotating over many orbits due to slip in a plain belt drive.

Alternatively the same action can be achieved by two matched sets of crown and pinion (or bevel) gears, linked by a pinion shaft along the length of the arm.  One crown gear fixed at the center of arm rotation, the other on the object.   Some care must be taken to get the gear set orientations correct - if the crown gears face the same way, the pinions must be on the same side of them relative to the center of arm rotation, but if they face opposite ways the pinions must be on opposite sides of them, e.g. the side nearest the other.

Another approach uses two identical plain gears on the object and fixed at the center + any odd number of idler gears of any size(s) to transfer the orientation out along the arm.

An electrical approach would be a pair of Selsyns, one at the center, mounted to the arm with its shaft fixed, and the other at the end of the arm.

[voltsandjolts]

If not in horizontal plane, use gravity.

[Mechatrommer]

so far as understood... more detailed drawing and connections will require professional service... 

[Doctorandus_P]

Two pulleys with a timing belt would also by my first choice (without knowing further details).

Another option is to look at the mechanisms used in those old drawing tables from before CAD moved those into history books. Those mechanisms keep the rules in the same orientation while still allowing movement over the whole drawing area. Those have been built with parallelogram mechanisms, with linear guides and maybe other methods too.

[Mechatrommer]

Another option is to look at the mechanisms used in those old drawing tables from before CAD moved those into history books. Those mechanisms keep the rules in the same orientation while still allowing movement over the whole drawing area. Those have been built with parallelogram mechanisms, with linear guides and maybe other methods too.

https://en.wikipedia.org/wiki/Drawing_board

not that i can see be implemented easily from a single prime mover... you can find it almost everywhere on our electronics benches..

[Infraviolet]

Thank yuo, that gives me a fair few angles to try from.

[LM21]

There are XY moves (or whatever they called)which can do that. Like in 3D printers and such.

[Neomys Sapiens]

For rotation input on A, object on joint D should do what you ask, I think.

(Reference: Mechanisms and mechanical devices sourcebook, Sclater/Chironis, 4th ed., McGraw-Hill 2007)

[beanflying]

I am having flashbacks  Mechanics of Machines, Hannah and Stephens - worth adding a copy to your library seems to be plenty secondhand out there.