A Saga of a Design Gone Wrong...Well Sort of...

This page, with its sarcastic title, is about a clock I started long before I understood anything about horology. I started off with a set of plans for the Thomas Clock and after looking at it thought it was rather plain. I soon set off on my own design with the idea of using the basic wheel work and layout etc from the plans. I had heard that using solid wood for the wheels was a bad idea as the wheels will go oval and cup with varying humidity. I was also not keen on the idea of wood on wood bearings. To solve the first issue I planned to make the wheels from Baltic birch plywood. To solve the latter problem I planed to use ball bearings which would be covered. I also decided that I would use lantern pinions in the time train (not the motion works) so as to further minimize frictional effects.

It wasn't long into the design that I decided I wanted to add extra features. I wanted moon indication and the best way I could see to achieve that was by using having a moon ball rotate behind an open disk. The 2" moon ball was to be half light maple and half dark walnut To propel it I decide that a 57 tooth gear on the hour arbor driving a 59 tooth wheel to the side. The 59 tooth wheel would have a crank shaft that drove a connecting rod that went up to the top of the clock where it attached to a bell crank. The other arm of the bell crank was attached to a horizontal arm with a hooked end on it. The hook laid on top of a 57 tooth ratchet wheel under, and connected to, the moon. Hence the crankshaft oscillated a little slower than once every twelve hours and via the rod linkages pulled the 57 tooth moon wheel through 1 tooth. Since the 57 tooth wheel on the hour arbor rotates twice a day the resulting ratio is 2 * 57/59 * 1/57 = a 29.5 day moon period.The period of the connecting rod going up and down (once per 12.42 hours) was the same as the tides so I decided that an oscillating tide plate (ala Fergeson) could be achieved. I decided that I could do that idea one better and include Spring and Neap tides. Here's where it gets a little complicated.

To the up and down connecting rod is attached a sideways "see-saw" rod laying over a pivot. Attached at the other end of the see-saw rod is the tide plate - an oval piece of maple with walnut veneer covering the bottom half. The top of the walnut veneer is shaped like waves. Like the moon this can be seen behind an oval plate with a hole in it (the tide plate can be seen above the face of the clock). Hence as the crank moves up and down every 12.42 hours the sideways rod see-saws over its pivot and lifts the tide plate up and down. The pivot however is not a fixed pin. It is a long rod pivoted horizontally at the back of the clock and protrudes out the front of the clock. Midway it has a pin pointing straight up that rides in an oval slot in the bottom of the moon ratchet wheel. Hence as the moon goes around the oval slot causes the pivot rod to move back and forth sideways. This has the effect of changing the leverage of the horizontal see-saw rod that causes the up and down motion of the tide plate i.e. twice in a lunar cycle there is maximum leverage and the tide plate moves up and down a little (Neap Tides). At two other other times in the lunar cycle the leverage is minimal and the tide plate oscillates a lot (Spring Tides). The back and forth pivot has an indicator arm on it which points out whether spring or Neap tides are in effect. The 59 tooth wheel mounted to the side which drives the crank shaft is marked with the hours to the next high tide. The Moon also has a wheel below it which indicates which day of the lunar cycle it is.

I also added another wheel, a spiral and linkage at the back of the clock which is used to indicate the day of the week. This is indicated at the top of the clock on the top of the moon face. Finally another spiral was added to the hour arbor at the back of the clock. This spiral causes a hammer to be pivoted to the side only to be released on the hour. The hammer falls and hits a wooden gong to strike the hour. All these extra indications were added at the slow end of the time train so as to minimize the frictional effects on the running of the clock.

Finally I wanted a skeletal frame and it was designed so as to support all the above features. I extended the skeletal frame to include a skeletal stand that had flowing lines that harmonized with the rest.

After designing the clock it was time to execute the build. Rather than cut wheel teeth on a bandsaw or scroll saw I made my own index wheel by plotting out large divided circles (large so as to minimize angular error). I printed this out and glued it to an aluminum disk which I then drilled out. This was then mounted to an arbor and secured to a frame (see photo below).

In order to cut gears I purchase some 14DP involute cutters and an arbor to mount them in my mill/drill machine. A bunch of blanks (incl. some sacrificial ones on the end) were mounted to the arbor and by moving the mill table back and forth while indexing the wheels were cut (again see photo above). The frame and most of the mirad of parts were cut out on a scroll saw. The lenticular pendulum bob was turned on a lathe. Having made all of this I counter bored the frame for the bearings using forstner bits. Here is where things went wrong. The (borrowed) forstner bits didn't bore snug fitting holes but rather oversize holes. This meant the bearings were going to be loose. While I was pondering a way out of this disaster I also decided that my gears, while functionally very good, were aesthetically mediocre. I didn't in fact like the look of the "plys" in the Baltic birch ply. They were too "plywoody" and did not have a pure wood look I was after. Additionally the outer laminations on some of the teeth were also prone to delamination. I also came to think the whole idea of using ball bearings was wrong. I've since seen many better ways of doing bearings. Between these outcomings (and some other thoughts about the core design) I stopped work on the clock. I have since left it in more or less the stage of completion shown below.

So there it is. In many ways it was failure yet in other ways it was a great success. If you work the individual bits it appears all things would or could work very well. It also taught me many basic concepts of clock making. Unfortunaly given the look of the gears and the oversized holes in the frame I can't see how to rescue it to a level I'd be satisfied with without redoing most of it.

Oh well, we all make mistakes. They key is to learn from them!

Back to

Back