Begin dial bezel work; begin debugging and finishing of calendar module -
June, July 2020
In this installment Buchanan begins to make the bezels for the
multitude of dials on the clock. At the same time the calendar module is
disassembled and the difficult job of debugging and then finishing begins.
The calendar module is possibly the most challenging to have constructed and
debug. Although it has about the same number of parts as the tellurian and
Sun/Moon rise-set modules, it's design is completely novel and shares little
in configuration or fabrication with conventional clockwork mechanisms.
The first thing that must be done before beginning the bezel work is to take
an inventory of the dials. These are only the main enamel dial works. There
are many subsidiary and adjustment dials that are made of engraved brass and
then finished by a French silvering process to give the surfaces a silver,
crystalline look something akin to snow.
The entire master artwork for the dials was done by Buchanan and the dials
fabricated in China through Bob Crowder's ProClocks company. Bob's personnel
produced a product that was perfect for our project. This effort
began in February 2008 through June 2013. During this time some changes were
made and dials re-ordered. There are photos of the dials being created and
the artists who created them in a four-part series here:
Part 1,
Part 2,
Part 3, and
Part 4. In
November of 2018 we ordered a replacement orrery dial to add color to the
zodiacal figures to better coordinate with the tellurian dial. This was the
third design revision to this dial and can be seen here.
Below are the polychromatic dials. The multicolored figures and scenes
required several firings as each color must be applied and fired separately.
There is always the risk of a dial cracking or warping at the red-hot
temperatures of the kiln, and this is made even more difficult when making a
dial ring as opposed to a solid dial disk.
Tellurion dial, 6", (15cm)
Orrery dial, 11.5" (29cm) in diameter.
Planisphere dial. This was the most difficult to produce. The star fields and
milky way required thousands of dots, and the zodiacal names were all done
in a calligraphic font. The multicolored scene required several firings as
each color must be applied and fired separately. The blue gradation from
light in the center to dark as it approaches the perimeter required special
skills to mix the paint in differing quantity as the artist moves from center
to edge. One must remember that the color of the enamel that the artist sees
out of the bottle is not the color that is revealed after hours in the kiln.
So he must be able to know exactly how the mixture will look even though he
cannot see it when applying it. One might wonder why the zodiac figures are
of the same color when they are much more complexly colored in the tellurion
and orrery dials, the answer is that the dark blue background precluded
this.
Monochromatic dials.
First photo: names for the moon anomalies, pull-repeat button (2), winding
duration (4), pendulum beat plates (2). Second photo: Twenty-four hour dial,
thermometer, equation setting calendar, strike control.
First photo: day of week, date, leap year. Second photo: Mean time and
separate sidereal hour and minutes nested dial set. These dials presented
some difficulties as to fit, see below; the outer mean time dial is
stationary while the inner two dials rotate counterclockwise to the sidereal
hours and minutes. One dial has been left off, we will see it later in this
installment.
In the four photos above, Buchanan has made a stepped gauge from Corian to measure the dial's
perimeters and allow
for the metal bezels needed to fit between them. Each dial must be able to
comfortably fit around the gauge. There were some problems in the fit and
Buchanan had to do some very careful grinding around the perimeters of the
fragile and brittle enamel dials to get them right. Here there are three dials that must nest perfectly
and allow for the thin brass bezels between. I had always had some
reservations about this especially since over the years the outer main dial
had been remade and the actual enamel dials for the sidereal minutes and
hours since being completed in May 2013 were never all mated together
between the three dials until now. But in the end they were all made to fit perfectly.
The bezel blanks after being stress-relieved. This is a procedure through
which the brass blanks are first put into a furnace to heat the metal
removing whatever stresses may have still been present from the casting and
rolling operations performed at the the foundry. This process is used
whenever there will be a final part that is machined into a rather long, thin
part. For example, the base frame members are long in comparison their cross
sections and were all heat treated, here the blanks are machined into thin
bezels. If the metal is not stress-relieved, the thin fabrications may begin
to curl as it is machined.
Buchanan
writes:
I have the Tellurian and Calendar dial turned. I have thickened them
slightly to surround the outside of the mechanism front frame.
A very nice idea, another special
touch that makes the overall presentation so special.
The slightly resized dials required a new pair of sidereal mounting rings,
left photo, the beginnings of the bezel ring of the mean time dial, right.
Here Buchanan uses tooling to create the bezel rings decorative profile by
hand, next the completed ring before removal from the lathe head. See video
below.
The completed outer and inner bezels for the mean solar time dial. Later the
entire set of bezels will be plated in gold.
The build up of the sidereal dials, the inner dial for the minutes, left and the
outer for the hours, right.
The complete buildup of the three dials, composing the mean solar time and
sidereal dials. The equation of time will also be read off the mean time
dial.
To give the illusion
that each dial has an inner and outer bezel, the outermost mean time dial
has the standard thicker, milled outer bezel and thin inner bezel. The two inner dial rings each have a thin bezel on
their inner rims only but when installed share the next dial’s inner ring to
give the impression that they too have a bezel on their outer perimeters.
When done properly it looks better than if one tried to give each ring a
full set, as it would look too thick. This design retains a delicacy and is
necessary as the enamel dials are too close a fit to do independent dial rim
pairs. Notice in the first photo how Buchanan achieved a nice, flat
surface alignment with the multiple dials.
Video demonstration of the completed nested dial set for the mean
time, and sidereal time. Notice the fine tolerances between the dials.
Another thing to notice is how nearly perfectly circular they must be to
pull off this effect. This could only be obtained by Buchanan's additional
grinding work on the dial perimeters and accurately applied bezel rims,
something that is usually not necessary in a standard, stationary dial
application.
The tellurion dial with its inner and outer bezels.
Look
carefully at the picture to the right. There
are several areas where frames must be
countersunk behind the dial bezel. Here there are
two curvilinear arms which are just barely
apart from each other as they pass the
perimeter of the bezel rim, but instead of taking
the easy way of making one opening for both
frame arms, the tiny area between them is
left intact resulting in two openings that
fit the pair perfectly.
The photo
above shows another example of
countersunk frames.
The upper calendar bezel is now complete, below the three dial mounting
plates for the mean solar and sidereal time dials..
Buchanan
writes:
I managed to fit the sun moon plaques onto the Bezel. I put a curved bend
onto the plaques so that they match more nearly the profile of the bezel and
I think it is a good improvement. I made a curved former and pressed the
plaque into the bend with a rubber block. I had to increase the curve a few
times to counteract the spring back but in the end they were pretty good. I
also had to carve the corners of the pockets square by hand to fit the
square corners of the plaques.
The plaques fitted into the Sun/Moon rise-set bezel
rim.
The bezels are now complete for the calendar work. The dial to the right,
the month and seasons is being remade now to correct for the Summer and
Winter upside down lettering.
Dials and outer bezel mounted to calendar frame.
Calendar frame mounted to clock. Note the close fit between upper
semi-circular bezel and the lower circular bezel.
--------ooo000ooo--------
Buchanan now turns toward the debugging and finishing of the perpetual,
third-order, reversible calendar module.
In this video Buchanan disassembles the calendar module.
Partial calendar module de-bugging list
1Pin trip cross
lever
2Pin S trip lever
3Fix / clamp
reader lever springs
4Shorten date
lever.
5Stop on date
trip lever
6End shake
control on 400 year cam assembly.
7Fix year
counter detent bearing
8End play on
leap year main arbour
9Straighten date
trip coil spring, will be done in the
assembled calendar
10Pin year
counter in feed bevel to arbour.
11Adjust year
counter.
12Replace clear
jewels with Red jewels to be done after
testing and polishing. Half done
13Pin cams to cam
shaft. 9 pins.
14Spring drags on
all dial arbours
15Pin balance weight.
16 Missing screw in 4 year cam
These are the parts of
the calendar trip mechanism, it is what drives the module’s computational
cycle once daily and is a sub module of the calendar mechanism.
These three photos are of the drive and logic cam
components of the calendar trip mechanism.
The first photo shows the timing camshaft (logic cams). This is the
coordinating component of the system and synchronizes the movements of the
various levers that move the other parts of the calendar. It rotates
one-half turn per cycle and its rotation is driven by the spiral spring. It
acts as the ‘clock’ found in mechanical analog computers to time the rest of
the components in the computer.
The fly blades in the first photo are part of a governor to control the
spring release.
Various wheels and their build-up.
Month index wheel, left and day of week index wheel, right.
These two photos show the month and day of
the week indexing wheels. Again one can only marvel at the delicacy,
artistry and craftsmanship that it takes to create these parts. All the more
so when one realizes that the same function could have been accomplished
with a tenth of the effort using conventional designs.
These three photos show
the year trip function linkage and is one of the more complex lever
assemblies in the calendar and provides the output to digital
display for the year.
The year trip linkage is shown here just before the assembly of the
calendar module in July of 2015. There are three layers of linkages each
subsequent layer relying on information from the former, their order is:
Day/Date, Month and Year. This photo shows the year linkage and how it operates
as an analog computer component, in this case a dual AND gate.
This photo shows the
various components that connect to this linkage. Notice the logic cams in
the lower left hand corner. It is the component that was shown earlier in
this installment with the calendar trip mechanism. The year linkage receives
input from all the components in the calendar that will decide when the
next year is to be displayed. In a perpetual calendar that condition follows
rules to keep the man-made calendar in step with the seasons as time passes.
That means adjustments need to be introduced and the first is the leap year
every four years, (a first-order perpetual calendar), then the cancellation of the leap year once every 100
years, (second-order), and the re-introduction of that adjustment every 400 years,
(third-order). To accomplish this the calendar must become a simple analog computer. To do these
calculations in forward and reverse the calendar has the four components
needed to qualify as a computer: Input, Processing, Output and Storage.
When one hears of a clock having a perpetual calendar function it means that
it is only of a first-order capability.
How could anyone
looking at these parts think that they are part of a clock?