dial drive wheels. Begin the reversible, perpetual calendar
calculator module - January 2015
The crossing out of the
calendar dial drive wheels is now complete. The smallest wheel, Buchanan
tells me, is as small as one can cut on the jeweler’s saw which he has used
to cut all of the flat stock on the clock so far. Smaller wheels than this
will have to be spoked out through a spark erosion (EDM) process.
These photos show the fabrication of the various pinions for the calendar
dial drive wheels. Note the matchstick for comparison in the fourth photo.
The solid disks have
small ‘V’ shaped groves which will act as does a star wheel in conventional
calendar work. These serve to make the dial hands jump from one date to the
next in discrete steps.
This design is made necessary because all of the dial drives are directly
geared together and not actuated by a pin engaging a conventional star
wheel. The last photo shows the tubes that will carry these count
The count wheels are
now spoked. When I first saw them I was pleasantly surprised in the way
Buchanan followed the contours of each ‘V’ groove rather than making the
interior of the rim a smooth circle. Once again his elegance shows through
in the extra efforts made. The third photo shows day detent wheel mounted to
its cannon arbor.
The first photo shows
the detent wheel with its cannon pinion mounted onto the arbor of the mating
begins to make the reversible, third-order
The second photo shows a wheel
pointed out by the pen that is fixed to the frame in the calendar mockup, it
will act as the ‘sun’ in a sun and planet epicyclical gear system. Next
Buchanan points to the position that this sun gear will occupy. There will
have to be a bridge built just above this drive wheel to fasten the gear to.
The last two photos show a preliminary view of how the calendar drive wheels
will look behind the dial cluster.
This is the sun gear
and it’s safe to say that this one will be the first wheel in the one
hundred or so made to date that will not be spoked out!
sun and planet gearing is shown in these photos.
Next the date detent wheel is shown next to the same wheel in the mockup.
That squared plastic tooth wheel is three times the size of the actual one
The last photo is an overview. The top gear covering the date detent
disk is covering three layers of gears below. The
largest is 90 teeth and just under 5/8 inch, (1.5 cm) and the smallest is 35
teeth and ¼ inch, (0.5 cm); module 0.18.
Here Buchanan makes
what may be the smallest gear in this project at only 1/8”, (3 mm). The
first two photos show the setup on the mill. Next a close up of the cutter
and pinion before cutting. Next a few photos to show the scale of that part,
especially nested between the teeth of the largest wheel in this project, the main going barrel.
Last a 20x magnification of the wheel teeth during the depthing operation.
Here a small bridge,
similar in design to that found on a tourbillon is fabricated. The purpose of this is
to hold the small attached gear as seen in the first two photos. The bridge
is in place in the third photo and that small gear is held in place, facing
upward and floating just above the large wheel below.
Here the tiny gear gets
installed and is shown with its chaton. The next photo shows a wheel that is
mounted to the same arbor as that gear from below. Third photo shows the
wheel from above mating to the pinion. The last photo shows the completed
assembly with the gear (red arrow) and the other fixed gear that was
described in the prior installment, (yellow circle).
The first two cams, the
four and twenty year correction cams, are now being fitted with their
associated drive gear.
Buchanan had to make
five tries to get this nine toothed pinion. Adjusting the cutter to the
center of the pinion blank is critical as an error of only 1000th
of an inch spoils the part. The module for this part is 0.15, the diameter
is 0.63 thousands of an inch or 1.6 mm. Next are the gears for the 20:1 leap
year cam. Tooth counts are 33:66 and 9:90.
We have now descended
to the smallest scale that will be encountered in this project. The three
tiny screws are used to attach the wheel in the next photo to its collet. At
this size it is no longer possible to spoke a wheel using a
conventional jeweler’s saw. The wheel thickness is less than the space
between even the finest teeth in a saw blade. These are now at the scale of
pocket watch work. The conventional way watch wheels are spoked is
either through stamping using tool and die methods or through the use of
extremely accurate computer controlled micro-milling machines. These methods
are only practical in a full scale manufactory producing parts in volume.
Here we are making a one-off part.
So Buchanan turns to EDM technology.
The third photo shows
how the wheels are cut out using an EDM machine. The wheel blank is
submerged in an oil bath while a copper electrode, above, is lowered to
contact that wheel. The electricity then vaporizes the brass blank where the
electrode is in contact. The design is completed through repeated contact
until the desired spoke pattern is formed. The fourth photo shows a copper
blank cut to the spoke profile and it is the positive die that electrically
erodes the brass wheel blank to produce the spoke pattern. Next a set of
copper dies for the various wheels in the perpetual calendar module and
finally some of the completed wheels. This method is slow, but once set up
requires little attention. The copper dies are made from relatively soft
metal and the pattern easily machined. We have used the EDM process before
to make some special conical holes in the main frame pillars but this is the
first time it was used to make individual parts.
Much later in the
project we will use this method to create a fine matte decorative finish in some of the
frame pillar parts.
A small cock is now
fabricated. Note the tweezers for comparison. Even at this scale two locating
pins are used as is best practice for clock and watch making. The third
photo shows the pivot being drilled with a very tiny drill.
The rough blank of the cock is now in place.
Filing buttons are used
to give the decorative contours needed to bring about the final contours of
the cock. And yes, we are jewelling the pivot. Every pivot of the estimated
1000 to be completed is either jeweled or equipped with roller bearings with
about two-thirds being jeweled.
Notice the beautifully
machined pillar that supports a steel collar. It is only a few millimeter
high. The next photo shows how small this part is as well as the fact that
the entire perpetual calendar module is on the order of a conventional
The attention to detail is penultimate.
Here we have the Geneva
cam which will drive the 100 year cam in 20 year increments, hence the five
The various parts to
the 100 year cam are shown. Note the tiny square holes in the center photo.
Here is the set up the
fabrication of the small square holes and square arbors. The second photo
shows one of the very small square posts at 20x magnification.
This photo shows how deep the calculator will be and the reason why
Buchanan had to have a the clearance which would otherwise seem unnecessary
for the rest of the calendar drive wheels behind the dial cluster.