Eric Watson, Saddleworth, England, 1986 

The clockwork section

Begin disassembly. In the photo, right, the remontoire is the drum and power spring both mounted above the escape wheel. The drum contains a small weight is a pool of dampening oil. The spring is what powers the escape wheel and therefore the balance wheel.

The remontoire supplies a consistent amount of energy independent from the force stored in the spring barrel. The design relies on the principle of inertia and is a solution that is equally ingenious as it is simple. Based on a 1943 invention by Henry Jeanneret, it ensures that the escape wheel at the moment of release does not get its force directly from the spring barrel, but indirectly from a spiraled hairspring that acts as a buffer. It is regularly wound via the gear train as soon as the escape wheel is stopped again. The pinion bears a small weight (seen here) whose inertia guarantees that the entire gear train only starts moving with a certain delay after the release of the escape wheel. At the moment of its release the escape wheel, is energized by the spiral spring, immediately takes off, only being followed by its pinion when it has already stopped again.

This allows the powerful mainspring to divert most of its energy to the orrery. Without the remontoire it would be very difficult for the clockwork to directly power the rotation of the heavy main armature as well as the over eighty wheels within the orrery.

The fork in the left photo reaches down to a spoke of the escape wheel, right and is what delivers the power needed to rewind the spring. The amount is so small and delivered so quickly that any error that is usually found in a spring remontoire is vanishingly small. As found this fork was located in the large area between the spokes, the fork being empty so it could not efficiently do the job as intended where the fork straddles a spoke of the escapement wheel.

The first photo shows the general haze and grime that covered the upper side of all of the orrery components. The oil leakage from the mainspring grease is evident on the plate, right. Fortunately the clock plates are all gold plated so the damage was minimal as gold is virtually inert to corrosion.

The grease leak was catastrophic. The original owner noticed the leakage less than a year after receipt. There was nothing to be done except to keep cleaning it with a Q-tip and some small amount of solvent. But there is no doubt this would have to be done carefully to avoid a smudgy mess on the surrounding plate surface. Either the owner got tired of cleaning, or the next owner failed to do so because the end result is obvious.

Here we see the problem. Watson had simply stuffed WAY TOO MUCH grease into the spring. Still it is curious that the grease appears to be stable, while the leak looks as if it liquefied to be able to seep between the narrow gap between the spring arbor and the barrel cap hole. Could the rubbing action of the coils have caused it to liquefy? I doubt the obvious which would be heat, since it started leaking in the original owner's home which one would assume to be at a reasonable temperature. The second photo is the setup on my lathe to remove the spring from the barrel and properly clean and lubricate.

The drive clutch which allows the demo crank to seamlessly take over the orrery for demonstrations and the automatically clutch to the clock drive when the demo crank no longer applies a rotational force.

The clockwork portion of the orrery is ready for repair and cleaning.

One of the first things to deal with was a binding between the second and third wheels in the train. There were two problems here. The first is seen in the pair of photos where the third wheel is catching on the second wheel collet, this is about a two thousands of an inch error. Too small to correct by straightening the second wheel, which would be problematic as it rides upon a friction clutch to allow for the adjustment of the clock time preventing such an accurate leveling as would be possible if it were fixed conventionally to a collet. The second problem contributes to this and that is that the second wheel had no end shake whatsoever. Even with these issues the orrery did run before disassembly, the mainspring being very powerful and Watson's clever use of gear ratios to provide ample power. Additionally I found in the trains several areas where as the wheels meshed they would come upon a tight spot between the teeth, some slight error in the tooth profiles. With some experimentation in the positioning of the wheels I found the 'sweet spots' where the erroneous teeth would not meet. I then marked the mating wheels so the next person would not have to play this guessing game. Unlike normal clockwork, in the orrery section most of the wheels mesh directly with each other, not through a wheel and pinion arrangement. This is why this problem can creep up more often than when the mesh is between wheels and pinions.

The first photo shows as found, the ball race on the top of the second wheel had already been dislodged by the arbor with no end shake. My solution was to not risk forcing the ball race further out. Instead I made a shim of eight thou to go between the plate and pillar directly opposite the second wheel ball race.

The wheel shim washer was thinned down to five thou. This subtracted from the shim gave me an extra three thou to work with. This solved the issue. There were another two areas in the orrery trains where wheel teeth were not correctly aligned, usually because there was too much play between the wheel and the screw that held it. Bear in mind that in a few cases these wheels are two inches or so in diameter, so any play at the center leads to a great offset at the wheel perimeter where the teeth are located. The tolerances in these areas were not as tight as in this case.

Overall the orrery clockwork cleaned up beautifully thanks to the gold plating. I might add that in some very small areas where the plate has been worn, I could see that it was a very thick coat, more like a gold fill than a gold plate.

Notice the remontoire output fork straddling the escape wheel's spoke and the very slight jarring which indicates the rewinding of the remontoire spring to drive the balance wheel.