View Full Version : Has anyone used a Sine/Cosine mechanical drive in a CNC?
08-30-2011, 07:15 AM
Ok... a little outside thinking after the EXPO last weekend and looking at the "MakerBot Thing-O-Matic 3D printer".... Which was very cool by the way !! Thanks wjsteele !!
Not that a belt drive or chain drive, or even a cable drive is all that tough to implement... I wondered if a Sine/Cosine drive had ever been used (I mean aside from moving a train down the track.) and most importantly if it might be even easier to construct.
I put together something very crude that I could at least hold in my hand with my daughters K'Nex, and although I implemented Sine and Cosine it was stable using four points with just a Sine drive or just a Cosine drive. If I were using only two points ( such as the X or Y drive) , I would definitely consider using both.
You-Tube link: (Proof of concept)
08-30-2011, 12:06 PM
Can you explain your idea a little further?
I can see this implemented two ways, the first way is just like you built with the print surface on any of the levels of your prototype... but, you wouldn't get the range of motion needed (only prints circles!)... and the second, you could use the side as the print surface's axis drive, you'd be loading the steppers differently depending on where on the sin curve the pivot point was. The problem with that is you'd have to compensate for it during your print or you'll wind up with artifacts in the print (which would be easy to find... they're obviously going to be along that sin curve... but none the less, they'd still be there.)
The beauty of the X/Y/Z arrangement is that no matter where the print head is, the loading on the steppers and drive system is the same.
08-30-2011, 03:48 PM
"but, you wouldn't get the range of motion needed (only prints circles!)" - No, I didn't see it that way at all as far as printing in circles.... this would simply be an alternate way to take the place of the current belt drive for the Z-Axis. The circular motion you mention connects directly to the four all-thread screws to move the platform up and down... not in circles. :-). I don't think the quality would be compromised, and especially since it is on the Z-axis it wouldn't see the activity level that the X and Y axis both experience.
Note: The exaggeration of motion is more because of what I had to work with to show as an example... if you lessened the length of the moment arm there would be a lot less circular motion, but with the same turning effect for each all-thread rod.
08-30-2011, 04:51 PM
Ah... you're just talking about the Z Axis. I was thinking about it moving the build platform itself. I got you... yep... that should work just fine... interesting concept.
08-30-2011, 08:36 PM
Is this helpful?
08-30-2011, 08:58 PM
I'll have to look at that in more detail, but I think that paper talks about moving the design plate in circular motions.
The method that I posted is a way to transfer rotational energy/motion from point A via sine and cosine push/pull rods and reconstruct the rotational energy/motion when it arrives at point B. The fact that the push/pull rods happen to make a square frame is more for convenience and perhaps an illusion, but it could be a triangle, hexagon, whatever.
The circular motion that you see was only intended to rotate the all-thread uniformly across all four points without the need to use a belt or chain driving mechanism. Just a light non weight bearing frame that connects them all together.