Bearing & Hardware

Showing 1-1 of 1 messages
Bearing & Hardware Christopher Boynton 2/14/12 8:02 AM
Hello Clock-makers,

I'd like to get a list together of all bearings and other hardware that we might need for the clock so I can start rounding it up.  I've been following the discussions so if you could just confirm the specifics to make sure we get everything we need in the right sizes that would be very helpful. 

Bearings? brass tube?  Just tell me what you need and I'll make sure we have it in the shop for next time.

Clock On!

On Mon, Feb 13, 2012 at 5:31 PM, Mathieu Glachant <> wrote:
Looks like you were totally right on the arms getting damaged if the inner rings hits them too close to their attachment point on the outer ring.

Surveyor printed the second design out, and posted a picture in the comments showing that the ABS whitened under the stress.

I'm going to have to add some more arm length, perhaps with a solid spur under it to provide support  to the inner ring...

I'll doodle something on paper to show what I mean, since OpenSCAD doesn't run on this laptop.

On Sun, Feb 12, 2012 at 3:48 PM, <> wrote:
Neat idea!

It looks like you're going to run into issues with the ratchet arm flex. The inner wheel begins applying its force at the bottom of the arm, meaning that it will need to do all its flexing right there at the base of the arm as opposed to being able to flex across the entire arm. You could possibly undercut the ratchet arms a bit or possibly even punch a circle at the base of the arm on the side towards the center of the circle (see the black dots in the pic below). That may introduce enough flexibility into the arm.

Alternately either increasing or thickening the total number of arms or taking a two tier approach should increase the axial load bearing capabilities. Though in both cases the ratcheting will become more difficult.

Here's a quick sample of breaking things into two tiers. Note, you can't really see it, but the arms are separated from the green disc by about 0.5mm in this picture. It emulates what you've got in one tier, while maintaining a high ratchet teeth ratio. The print is trickier and would need to be done upside down (with the arms on the build surface) and with support material. There are many variants of this with the bottom disc having various parts cut-out or even possibly replaced with triangular arms, etc...

I noticed that the ratchet drum you're using seems to be based on the mcad gears lib and the teeth are sparse with the ratchet arm faces not being in full contact with the ratchet wheel when bearing load. I believe this may make a difference in the load bearing capabilities as well.

Hope it helps!

On 02/12/2012 02:31 PM, Mathieu Glachant wrote:
I just published a derivative of RustedRobot's fine ratchet design, which is intended to improve the axial load bearing capabilities, removing the need for a complicated and thick part with spokes that reach over the inner ring to reach the shaft.

It works (hopefully, anyway) by having the outer ring resting directly on the inner one, with the ratchet arms only needing to stop the rotation, not carry the axial load which is now transmitted directly, ring to ring. This should allow us to integrate a ratcheting drum into any wheel without breaking the connection between large gear and pinion gear, adding more thickness than the minimum needed for the string, or  connecting the drum externally via some sort of physical API.

Plus, it's even prettier than the original ratchet, possibly. :-)


NOTICE: This email may contain information that is confidential or attorney-client privileged and may constitute inside information or trade secrets. The contents of this email are intended only for the recipient(s) listed above. If you are not the intended recipient, you are directed not to read, disclose, distribute or otherwise use this transmission. If you have received this email in error, please notify the sender immediately and delete the transmission. Delivery of this message is not intended to waive any applicable privileges.