Frictional winding

Due to the fact that the clock seems to need about 10 lb of weight, it seems like the endless chain mechanism I had intended to use probably won't work.  In particular, it basically requires doubling the drive weight, which doesn't seem like a good idea.  So instead, I've opted to try a simpler mechanism for winding.  Merely winding a few turns of cord around the drive pulley seems to give quite a bit of friction -- enough to hold the drive weight with a small amount of force on the other end. 

Specifically, it looks like if the drive weight is the 9 lb 8 oz from before, a 7.5 oz counterweight seems to hold it in place.  In this case, since there are two weights that need to be held clear of the pendulum and each other, I've used a pair of pulleys to run the cord.

This seems to work well enough, when the drive weight runs down, you can lift the drive weight and gently take up the slack with the counterweight.

Winding mechanism

The winding mechanism for the clock follows the design shown in Huygens' endless chain mechanism.  The pulley is attached to the center wheel arbor and can have a max diameter of 1.5".  It's a tight fit to the center wheel arbor and is secured with glue.  The ratchet wheel can have a diameter of 2.5" and is offset from the main axis, and its arbor is clear of the center wheel.  The ratchet wheel needs its own arbor and a click.  (I expected to need a spring to keep the click seated, but that appears to be unnecessary.)  Here are the wooden parts for the ratchet wheel (left) and pulley (right).

The ratchet wheel also has an arbor and bearings, though that's perhaps unnecessary, it seemed appropriate.

The ratchet wheel bearings are placed offset from the train, as shown below:

The pulleys are friction-fit onto their arbors.  Unlike all of the gears that I attached before (which required several attempts with the lathe to turn the arbors to the correct size, then undershooting and building back up with solder), I tried a different approach.  Simply grip the arbor in the lathe, and then use an empty drill chuck in the tailstock to push the part onto the arbor with the lathe off.  This worked very smoothly!

It took several iterations to set the position of the ratchet wheel's pulley along its arbor, so that there is enough clearance between the center wheel and the ratchet wheel.

The pulleys both have several parts glued together.  Here is the initial gluing setup for each.

I just used Elmer's wood glue, and weighted the pieces.

I attached the click to the middle plate with a small nail.

I had expected to need a spring to keep the click secured while winding.  However since I located the click near the top of the ratchet wheel, gravity is sufficient to keep it anchored.  This also makes it easy to release the click should I need to.

Here is the cord attached.  I'm using 1/8" nylon braided rope with two single sheave pulleys for the weights.  Even with two turns on each pulley, the cord seems to slip, but it is sufficient to demonstrate that the mechanism is correct.

 

(Update 2014-01-17) One thing that is also slipping is the center wheel on its arbor when a large torque is applied to the drive pulley.  To fix this, I applied a small amount of glue to attach the pulley to the center wheel.