Torsion clock pendulum project

[EricB]

The thing that got me started in model engineering was clock collecting. Somewhere along the way I figured I would need to learn to fix them too so I started with what I could do with only hand tools.

In 2010 I acquired my first torsion pendulum clock, an Atmos. That lead me to get a second torsion clock, a 400 Day or "Anniversary" clock. Over 130 clocks later brings me to this project.

In 1981 the August Schatz Co. celebrated their 100th anniversary in the clock business. To commemorate the event they issued 3000 numbered "Jubilee" clocks based on a design they had made in the mid 1880s. I already have one, but recently I found a second without the pendulum. I don't expect to find a loose pendulum any time soon so I decided to make one.

[EricB]

The two small disks on top of the large disk are for adjusting the time keeping. They run on a screw that has left and right hand threads on the opposite ends. That way turning the screw will move both weights in oposite directions, toward the outer edge of the large disk to slow the clock or toward the center post to speed it up.

I could probably make the screw but I had some parts to salvage from a clock made in the early 1880s.

[EricB]

I did not have the brass for the large disk, so I started with the center post.

Thinking I could save some effort, I reused the bottom nut from the 1880s parts. I found out today that was a bad choice. Anyway, I made threads to match the nut and turned the rest of the post to match my sample.

[EricB]

The slug for the large disk arrived, so I converted about 2/3 of it into swarf. What started as 750 grams is now just 222.

I use Sherline tools. This part was too big for the chuck and stalled the motor several times but I finally got through it. I turned a bit on the outer edge thinking I would use that to dial it in when I reversed it on a 4-jaw chuck. Then I just used the same 3 jaw. I later put it on the mandrel I made later and trued up the OD.

The finish on the top did not come out as nice as I would like. I had some trouble with chatter at the outer edge.

[EricB]

The next part was the center finial.

By now I've made quite a few finials for myself and other peoples clocks, normally in groups of three that need to match. Making just the one is much less taxing. The basic shape was roughed out with round nosed tools and I finished using gravers.

On this pendulum the finial is the nut that holds the gallery in place where most are just decorative. The bottom was threaded as was the top of the center post.

[EricB]

The top and bottom pieces that make up the gallery are probably the most complicated parts to make. The shapes are 35mm diameter circles that are cut in by three 20mm radii 120 degrees apart. The center distance between the part and the 20mm radius determines the final shape and size.

To make these I needed to sweep the radius and index the parts. I piled up a rotary table with an accessory plate and an indexing head. The parts where held on a mandrel made from the stub the finial was cut from.

I started with thicker stock just large enough of the parts and turned them into semi circles. After I trashed the first one I found this step was not needed.

I made three of these. The first was too small because I did not get the center distance right. I also did not have enough travel in my setup and could not complete the cuts. I used the scrap as my test piece for later operations.

I did get the curves and sizes right in the end.

[EricB]

The next steps for the gallery parts were reducing them to their final dimensions, drilling holes for the three steel pillars, and making the pillars.

The drill rod I used for the pillars was a few thou larger than the holes left by the drill so I sized the pillars with some sand paper.

The final work on the gallery was to drill and tap for the custom screws that hold it all together, and make all the screws.

[EricB]

The final machining on this project was to make the hook that attaches the pendulum to the clock. The sample I'm copying has the hook riveted in place. The original parts had the hooks threaded so I went that route. That way too I could fix all the finishes later.

I did not take any pictures of the process, but I started with the threaded end using my scrap part to get them to the correct length. From there I turned the OD and parted off a bit long. I threaded a mandrel to match the hook blank and finished the top shape.

From there I made the hook itself. I've made number of these so I have a special fixture. First the part is cross drilled for a pin, then the central slot is cut from the top down. The part is rotated 90 degrees in the fixture, and the fixture repositioned, then the side cut is made to complete the hook.

None of this build was very difficult, but there are not many people who do this kind of work.

[EricB]

Here is the final part in action. I said earlier the using the old nut was a bad idea. You can see the large disk wobbles because the threads in the nut are not square with its mating surface. I'll sort that out when I polish it up.

Thanks for looking.

Eric

<a href="https://www.youtube.com/watch?v=dUw8E3m2sXc" target="_blank">http://www.youtube.com/watch?v=dUw8E3m2sXc</a>

[Chipswitheverything]

Thanks for the write-up and the excellent photo sequence outlining this quite tricky horology job, interesting.  Dave

[Dave Otto]

Nice work Eric!

Dave

[John MacArthur]

Nice work on a finicky part!
Johnny

[Kim]

Very neatly done, Eric!   

I've always wanted to make a torsion clock like this someday!  Maybe that'll be my next project after my current clock... or at least on my to do list somewhere! 

Kim

[EricB]

Thanks for the comments!

I had thought about making different pendulums before but always managed to find the ones I needed.

I went ahead and cut new threads and made a new nut for the bottom. The old nut was somewhere between 2BA and 8-32 ASME so I was able to go with 8-32. The wobble is gone. I'm happy with how it came out.

Eric

[bent]

I like that adjustment device.

[EricB]

I like that adjustment device.

The person who purchased the patents for torsion clocks back in 1883 was a marine engineer. He adapted the adjusting screw idea to lifeboat davits and received a patent for that too.

Eric

[Zephyrin]

very nice work on this clock, fascinating mechanism !
In order for the spacing between the two discs to be dead symmetrical,
I wonder if the two opposing threads need to be exactly opposite each other, i.e., start at exactly the same distance on the axle. The same applies to the nuts on the disc supports; otherwise, it seems to me that there could be a maximum definite error of half a thread.

[EricB]

very nice work on this clock, fascinating mechanism !
In order for the spacing between the two discs to be dead symmetrical,
I wonder if the two opposing threads need to be exactly opposite each other, i.e., start at exactly the same distance on the axle. The same applies to the nuts on the disc supports; otherwise, it seems to me that there could be a maximum definite error of half a thread.

Thanks for the comment.

I like a good thought puzzle.

I'm thinking the only important dimension for the threads on each end of the adjusting screw is pitch. In this application, so long as the parts move in opposite directions by equal distance the angular errors in the threaded parts are irrelevant. In operation the pendulum oscillates at only 8 beats (cycles) per minute. Dead symmetrical spacing between the two disks isn't much of an issue at that rate because the variation in positions of their masses by a few threads isn't enough to throw the system out of balance.

Eric

[bent]

A case of perfection being the enemy of good enough?

edit to add:  If it were really important, one could an adjustment screw at the two mass attachment points for the ends of the differential threads, so you could set the spacing as close to equal about the rotation axis as possible.  Then make the differential screws run through a bimetallic rod, like some pendulums use, for temperature compensation...

Sorry, my mind sometimes take the journey down the rabbit hole to ridiculous extremes.

[EricB]

Sorry, my mind sometimes take the journey down the rabbit hole to ridiculous extremes.

Oh, I've been down that rabbit hole a time or two. At one point I was building a test fixture to see how different factors affected the pendulum's performance. What I found was the tiny rare earth magnet I used with the hall sensor to see the rotation was able to induce eddy currents into my aluminum fixture and stop the clock. One day I will start over with a wood frame and no metal near the magnet. One day...

It was fun to try though.

Eric