June marks the
completion of the machine and the journey now to be trod through testing and
ultimately the travel of thousands of miles to its final destination and
reconstruction. As this process unfolds, there will be a few additional bugs
and tweaks made as needed. This month an intermittent problem with the feed
pawl to the celestial train's Robin remontoire is addressed. Bias springs on
the strike flies needed to be adjusted as they were a bit too strong. The
keyways begin to begin to be located for the front glass.
The sub-base wood
surround is built. Other small adjustments are made. The dedication plate
and winding key
design is discussed.
Next month the trialing of the
machine for rate and accuracy begins.
The raw base material is shown above, the surface is veneered flame
The completed sub-base surround is finished, except for a brass trim piece
that will follow the perimeter of the bottom. The surround will enclose a
2.5", (6.5cm) thick aluminum base block that the clock will rest upon. The
aluminum base is needed to ensure that the clock's brass base will not
suffer damage from any uneven surface that the clock
might otherwise rest upon. The brass clock base is made from four major
components which could shift under stress. Attached to the clock's base are several
tall frame components that would magnify any errors in the base frame and
could result in the binding of the entire main wheel works.
Before the machine was converted from weight to spring drive in October of
steel stand served the function of the aluminum block. After the conversion
the stand was no longer needed for the weights to be suspended.
Shown above is my first initial idea for the dedication plate. I did not
have the capability of making a bare oval so there is a white square that
will eventually be oval. There is nowhere on the front side that a plate could go, so
the rear was the only place left with the space needed. Looking at the front, a
dedication plate even if it could go there, would be a blemish. Another
option would have been to replace the Latin inscription created in May 2018 to be around the
planisphere mask with the dedication, but that, in my opinion was not an
option since the phrase so nicely encapsulates what this machine represents
and compliments the stars represented on the planisphere dial.
This is just the initial design and information.
Buchanan had made something similar in the past for another client, shown
above. I like the 3D depth, especially the raised lettering in the oval
center that contrasts with the engraving within the perimeter.
Buchanan asked about the type of winding key to be made. I suggested a
curvilinear design similar to one shown from my collection, but with a couple
of ivy spurs like those found elsewhere in the frame design, and if there
was enough of the mammoth ivory left over from the creation of the tellurian
Earth globe back in September 2016, that this be used for the handle. I've always liked the heavy, yet silky, warm feel that ivory has in the hand. Over time a
develops from handling. I also suggested that Buchanan may want to have some
sort of engraving that will connect the key to the clock.
Correcting a bug with the Robin remontoire feed pawl.
I have a bug with the feed rod from the rear time remontoire to the Robyn
remontoire. I want to make it lighter as it is affecting the poising of the
Wagner carriage. It should be a simple fix. There is also 1 link in the
Robyn chain that very occasionally catches on one of the sprockets. I have
had a video camera on the clock most nights and one can pick up a single
delay as the relationships between each of the Wagner carriages and the
positions of the pulleys on the Robyn must remain consistent. It is very
easy to go through a whole night’s video in about half an hour and find one
small change. I can then replay it frame by frame to see exactly what
Having the machine under video surveillance is a good idea. It is a complex
mechanism and having many dials that record time or position lends well to
detection of any malfunction. Buchanan added a counterweight to lighten
where the pawl drives the Robin remontoire pulley (left photo above). The
original mass of 4.4 grams is eventually reduced to one.
With the very light feed roller force, when the Robin remontoir is released
and runs down, the two curved latching arms have to be stopped by the
unequally spaced release cam (black arrow). This is driven by
the roller feed pawl with the new light counter balanced rod. It seems as
though when the curved latching arm hits the cam, it knocks the feed wheel
on an extra step (yellow arrow). I had this problem may years ago
and I don’t quite recall whether it was because I had counterbalanced the
feed rod, like we have done now. Whatever the reason, there is a very simple
fix. All I have to do is to reduce the diameter of the unequal release cam.
This is about 5/16 inch diameter. I can easily reduce its diameter to about
half of that and, as this is an intermittent problem a 100 percent
improvement in the leverage ratio will fix it completely.
The diagram above shows the feed pawl acting upon the Robin remontoire
pulley. What Buchanan refers to as the unequal spaced release cam is what
allows the Robin remontoire fly fan governor to cycle at five uneven intervals. What this means
is that, in contrast to the main time train fly governors which cycle every
thirty seconds, this cycles at five differently timed intervals with a mean
time of once per minute. Since this is simply a timing of the release of the
remontoire to allow the celestial main train to rewind the remontoire, as
long as the rewinding is occurring within the one minute cycle on average,
there is no adverse effect on the accuracy of the mechanism. The concept of
unequal spacing of the time intervals was chosen to make the Robin fly
governor action somewhat unpredictable and therefore more interesting and
attention getting. An even and predictable pattern soon becomes ignored by
the the human mind, this will add an additional feature, particularly since
it is purposefully set deep within the machine, so one will at first notice
something shiny glittering behind the complexity demanding from the observer
a closer look.
So essentially this bug results in the feed pawl dropping occasionally two
steps rather than a single step. This will cause the take up pulley to run
too fast in relation to the feed pulley causing a lockup of the remontoire
from the slack in the chain being taken up. We have designed for this
eventuality because both the time and celestial train share power from a pair of
springs (formerly weights) that drive each train. The reason for this design has been explained,
but briefly the time train requires about ten times the power of the
celestial train. We did not want one weight to be so outsized from the next
weight over, thus the sharing arrangement. When the conversion was made from
weight to spring, the same constraints were present. A spring large enough
to drive the time train on its own would not have fit into the extant space
occupied by the original barrel used for the weight line.
Since these two trains share their springs I wanted to avoid the clock
stopping if the celestial train should fail. A complex clutch system
overrides the Robin remontoire in the event of a lockup, allowing the time
train to continue drawing power from the springs.
The new unequal cam, left, will have a smaller diameter than the old one,
right. This is illustrated in the drawing above by the black arrow. Instead
of a hollow cage with unequal openings, the same design will be cut into the
center post allowing for the diameter to be cut in half; lowering the force
the remontoire locking detent exerts on the cam.
Another small change is made as follows, Buchanan writes:
I am still working on the Tellurian dial. I have brought the Tellurian
forward by 2 mm as it was rather close to the enamel dial. This is just a
long term insurance policy. The back of the tellurian mounting post was a
bearing for the strike trip so I had to extend that bearing holder. This is
not all in place. I am going to have to machine a little of a section of the
Sun/mod dial bottom, the concave section. Fortunately I can easily unscrew
it without dismantling the Sun/moon dial.
I am constantly amazed how easily most components can be reached in this
clock. It is very seldom more difficult than removing a wheel on a chiming
clock and mostly very much easier. By removing one, or sometimes both
chatons, many wheels between the frames can be taken out. Almost anything
mounted on a dumb arbour is easily removed and often screws that appear
inaccessible can be reached from the other side of the clock with a long
This is the result of Buchanan's careful attention to design. One of these
being the extensive use of chatons within pillar frames rather than
conventional jewel pivots or ball bearings friction-fitted permanently into
a conventional plate. A chaton is a removable pivot, so when one or both are
removed the wheel arbor is completely free of the position it held between
the frames without the need to part those frames. This is but one example of
the many features designed into the machine to make such a complex mechanism
as easily serviceable as is possible given the intricacies involved.
The clock is connected for the first time to a Microset®
timer to begin test and rating. This will begin in earnest next month.
The Astro-skeleton is now ready for shakedown trialing.