Back Up Next

Debugging and timing trials continue, small tweaks are made - August 2021

This month the clock continues to be trialed. A few small tweaks are made to the planisphere setting knob, a pawl in the strike train, some improvements in the calendar module and a decision on the aluminum base.

The first item was a decision on how to treat the upper surface of the aluminum base, that portion upon which the clock will rest. Basically the two choices were black, so that it would match the black top rim of the wood surround, or  a polished surface giving a metallic, shiny and reflective surface.

The wood surround is what hides the aluminum sub base. The 2.5" (6.4 cm), aluminum sub base is needed to be sure the brass flat bed of the clock has an absolute flat and solid surface. The brass flat bed frame is made of several parts, and given the weight of the machine it  could flex a bit at the connecting points. This movement would be magnified by the tall main frames going out of alignment with respect to each other causing the wheel works to bind resulting in a catastrophic failure. Except when being moved by a special device supplied with the movement, it must always rest on this block. The block has three small feet so it cannot ever rock, no matter the condition of the table top.

The photos below alternate between the two choices.

This photo shows the unpainted base. It still has machining marks and will have a better surface for this choice. This is left hand, end view.

This photo shows the same area in black.

A view from the rear.

A rear view with the black surface.

A view from the front in silver.

A view with the black surface. One can see in this photo where the color of the wood surround comes into play with the painted sub base. But in the end, the natural silver of the aluminum base was chosen. It gave more reflected light throughout the machine.


Next the holes for the front glass keyways are drilled in a temporary Plexiglas that will serve as the front for the mock up case.

Seven holes are needed for the front, four for the winding springs, two for the celestial demonstration functions and one to set the main time dial. Look carefully at the four, larger holes, they have a nice brass ring insert! Even on a mock up one finds perfection.

Area within the yellow circle shows where Buchanan has created a grub screw to attach the calendar detent wheel. The petal shape, see photo below, is what makes the year flip instantly and stay in position. Prior to this it was a tight friction fit making adjusting more difficult should the need arise.

Another view of the screwdriver inserted into the grub screw behind the calendar detent wheel. Just above one can see the the jeweled detent that rides on the rim of the detent wheel. This wheel is actually a reused part from what was to be a part the escapement originally built in January 2009, but the design was eventually discarded for another. Nothing goes to waste.


Here Buchanan has added a clutch to the connection of the calendar module and the celestial drive. Again this makes adjustment easier if one needs to independently move the calendar module. The first photo shows the exploded view of the clutch, which is basically a small spring ring with a knurl knob for tension tightening.

The completed clutch assembly.


First photo shows the making of small screws used to attach the coordinate points plaques to the planisphere's spider mask. Yes the coordinates look to be backward, but this is the correct positioning for the planisphere and is NOT the result of the fact that the clock is being built "down under" in Australia.

This photo shows the reworked setting knob for the planisphere. Originally the knob with a very small number of teeth was permanently in mesh with the large wheel of 516 teeth. During the demonstration function this would cause the setting knob to spin very rapidly resulting in a bit of a rattle and was deemed to be distracting and a possible problem. The solution was to spring load the knob and its gear to remain out of mesh with the large wheel behind the planisphere dial. To use the setting knob one needs to push it down.


Buchanan wanted to make an improvement in the 'Strike/Silent' control spring; replacing it with a pawl. The parts are shown above. 

The pawl installed within the strike train to give better performance, yellow circle.

Buchanan writes: Here is this week’s run from last Friday afternoon, the 13th. I haven’t touched the clock since I wound it at the start of the run, it is driving all complications.

The long curved blue line is seconds lost so we are about two min slow. It will be interesting to see what the next few days bring and what happens when we wind it. I am pleased about the fact that the curve is constant.

I have just noticed that the amplitude is increasing and decreasing in phase with temperature.

I am not quite sure why the amplitude is slowly increasing. We do have a certain variation of about 15% in the spring strength over a week.  It may be the cause.

Buchanan writes: The clock ran down on Saturday afternoon after a full 8 days.

It ended up about 60 seconds slow for 8 days.

I rewound it without touching anything, and the same rate pattern started reappearing. You can see the stop period spike on the 21st.

The first week started with a losing rate and ended with a gaining rate.  When I rewound it, it  started to lose again.  This is a good sign, in that it is repeating the same rate profile.

There is obviously some feedback through the remontoir from the small change in mainspring torque.

You can now also see some small temperature error.

Total parts count to date, for the machine net is 7688, with the case 8009.

Back Up Next