Begin fabrication of drives to the sidereal, equation of time and calendar functions - September 2014                     

Back Up Next


The dial work shown is from a Tompion tall case clock I had seen when attending the Ward Francillon time symposium Time for Everyone, held November 7-9, 2013 in Pasadena, CA. This dial displays mean solar time on a stationary inner dial ring, and sidereal time on the outer dial that rotates counterclockwise twice yearly allowing sidereal time to be read directly off the same dial work and hands. One can see in the second photo the four rollers that hold the rotating outer dial ring. 


These photos show the design mockup for the revised sidereal time dial. After I had seen a Tompion clock using this design at the Francillon Ward NAWCC symposium last year, I asked that we use this instead of there being a separate dial for this function as was originally planned. The second photo better shows this concept. The inner dial rotates counterclockwise by just over 3 minutes a day with the main time dial remaining fixed. This allows one to read directly the difference between the two types of time with the same set of hands. It also allows one to actually read sidereal seconds. Of course during the elapse of one year the difference between regular time (mean solar time) and sidereal time equals one full day or twenty four hours. Since the inner dial is demarked into twelve hours, one must remember to add twelve hours to the time read off the dial during the second half of the year. This is why sidereal time when displayed on its own dial is usually denoted in twenty four rather than twelve hours. I think the extra mental calculation during the second half of the year is well worth the benefit of being able to read in real time the actual difference between the two types of time. Trying to mentally equate these between separate twelve and twenty four hour dials to me is more difficult. This dial set will also display the difference between ‘sun time’ and clock time also known as the equation of time. These dials and indicator hands will allow one to read directly from one dial set all three types of time to the second and comprehend the relationship between all three simultaneously.

The real downside of this is the breaking of symmetry between the time dial and the tellurian dial. The combined time and sidereal dial is thicker than the tellurian dial. There may be some action we can take with the bezel work on the tellurian dial to alleviate this issue. To be fair, perfect symmetry has already been broken with the setting dial for the equation of time kidney cam located below the time dial.

The diagram shows the difference between the sidereal and mean solar day. Briefly, sidereal time is a time scale that is based on the Earth's rate of rotation measured relative to the fixed star rather than the Sun. A mean sidereal day is 23 hours, 56 minutes, 4.0916 seconds (23.9344699 hours or just under four minutes every twenty four hours shorter than a solar day.

Drawn here is the sidereal, equation of time and calendar drive components. These originate from the center arbor under the escapement support structure, number 4, and flow four wheels to the left to split with two bevel wheels that each drive a worm gear, (two yellow arrows) the lower for the equation kidney to rotate once per year, and the other a worm to turn the sidereal time dial counterclockwise twice in one year. We could have used bevel wheels in place of the worms configured like those employed to drive the remontoire fly fans, but I wanted to introduce a worm and helical gear, components that have not yet been used in this project. The mating drive bevels will need to be of differing diameters and not the same as shown in this diagram. The second and third wheel from the center drive arbor will also contain winding squares that will be used in the celestial demonstration feature. Wheel number 2 has a winding square that will demonstrate the celestial functions at a rate on one turn for each 24 hour period. Wheel number 3 has a winding square and multiplies the turns per day by four. Number 1 has a winding square and demonstrates the interplay between solar and equation time. The red arrow points to a wheel with four pins which turns once in twenty four hours. These will drive the perpetual calendar.


First Buchanan makes the wheel train in plastic in order to check for fit within the existing movement. This also allows him to position these wheels in the most aesthetically pleasing way and to plan the frame work to support those wheels. This as well as the initial drawing of the wheel train against a photo-grey copy of the exiting movement is a standard operating procedure.

Now the wheels begin to be fabricated in metal. As has been done many times before, the rough wheel blanks are cut from brass stock, the center arbor holes are drilled and then the perimeter wheel teeth are cut. The fourth photo shows a pinion being cut with a fly cutter.


 Next the wheels in metal are fitted to the movement to check for functionality and fit before they go on to further finishing. These wheels represent the main drive before additional wheels are attached that branch out to the equation, sidereal and calendar functions. Next the outer plate blank is put into position. One can just see the inner plate behind the wheels in the first photo.

Here the additional wheels that branch off the three complications are made. Compare the fourth photo to the finished product later in this installment.


The outer plate shown earlier is now in place with a few of the additional wheels mounted outside. Notice the two winding squares with a key attached to the left most square. These are part of the three speed celestial train demonstration drive. Next another plate to go in front of those wheels shown in the prior photo as well as additional wheels that will continue toward the left.

The plate is now in position with holes drilled for positioning. Next sections are removed.


More parts are rough cut out of the plate as well as the new mating plate. We now have a separate “plate and spacer” subframe for the equation and sidereal drive wheels to be mounted within.

Next the arbors are finished to the familiar shape with tapered ends. Each of these are composed of a stainless steel arbor with hardened steel pivots fitted into each end. Next the rough frame is fitted out with jewels and then the arbors are inserted to check for fit.


The first photo shows the equation and sidereal drive wheels fitted to one of the plates. In just this one small sub-frame one sees as many wheels as in a simple conventional clock. Next the assembly is shown fitted to the rest of the machine.


Here is explained the three winding squares shown in the first photo. Number 1 demonstrates the interplay between solar and equation time. The two numbered 2 and 3 are the two speeds that one will have in demonstrating the entire celestial train along with the calendar giving a temporal reference to the entire demonstration. Number 2 will allow one to crank the square to represent one turn for each day. Number 3 is a divider equal to four revolutions for each day. Number 4, the next one over and directly below the escapement is the control for the two speed drive for the orrery. It also connects (locks) the celestial functions to the clock to operate them in real time and allows for disconnection of these for the demonstration mode. At the slow speed one will be able to run all celestial functions including the orrery by the first two cranks. The second faster speed is a 12:1 ratio and allows one to demonstrate the orrery disconnected from the rest of the clock since the outer planets have long orbital times with Saturn at over 17 years and this cannot be displayed with the calendar function providing a temporal reference.

To the far left is an additional square, number one, which has a key inserted. This demonstrates the equation of time and mean solar time showing the hands moving together and showing how the equation hand will precede and regress behind the minute hand by approximately fifteen minutes throughout the year. This is a stand-alone demonstration and is separate from the calendar function and the celestial train demonstration functions. This was necessary as even with the slow speed celestial demonstration speed at four turns per day, it would run much too fast for the solar/sidereal minute and hour hands. It is the same reasoning but for opposite conditions that the orrery has a separate high speed demonstration function separate from the celestial demonstration.  

The last photo shows the main time dial ring and the smaller dial representing the former design and location for a separate sidereal dial. In a fortuitous turn of events it turns out that this dial can be repurposed without any modification into a world time dial. We talked about having a dial located in front of the second wheel to the right of the equation wheel train. That wheel will turn once in twenty for hours. One idea is to have a ‘world time’ dial where various cities local time can be represented with a hand designated for each. Since the dial is rather small at about two inches, we could only fit five or so cities, say New York, London, Paris Moscow and Beijing. A pretty good idea as it adds another legitimate complication. We would need to balance this with another dial opposite to keep the overall symmetry of the dial layout across the movement. One solution could be to have another would time dial located at that spot, perhaps for the southern hemisphere. The fact that this dial is also delineated with two colors for each twelve hours on the twenty four hour dial gives both day and night indications for each city.

Now on to the screws needed for this part of the project, many which will be used to attach the wheels to their collets as well as other parts of the drive components. The first photo shows the raw rod stock for the screws made in the second. The next shows a close up of a saw blade used slit the screw heads Shown in the buckets are 160 screws 1.6 mm and 0.9 mm diameter.

Next the wheel collets are fabricated. After the blank is cut, the holes are drilled and tapped. The next two photos show the decorative machine work

The finished compound wheel set. Note the beautiful concave feature on the collet rim. Compare this to the wheel blanks that composed this piece in the sixteenth photo earlier on in this installment.

The first two photos show the front and rear of the two wheels with their winding squares responsible for the two speed demonstration of all of the celestial functions. Next the first of the remaining wheels in the drive train. The wheel characteristics become more delicate as one goes out from the main drive wheels.

Here we have the complete drive with all wheels planted on the rear plate, so what one sees is the wheel set from the rear. Note that if this were flipped around the large curved cut in the plate would match that of the main dial pictured in the diagram to the upper-left corner.


The first photo shows the existing post mounted to the time train frame post. Upon this are mounted the equation drive, kidney cam and setting dial. These parts are mounted onto a threaded cannon pinion shown in the second and third photo. The assembled parts are in the fourth. The knurled nut is the kidney cam clutch used to allow the kidney to be adjusted and then screwed down to lock it in place.  In the last photo the rear disk blank is the kidney drive wheel and the kidney cam will be fabricated from the front blank.



Here begins the sidereal time dial ring. First the hub upon which the dial and its drive will rotate is fabricated, yellow arrows. This is fitted around the existing triple set of cannon pinions and central arbor carrying the seconds hand. It is fortuitous that these were designed with enough length to accommodate the new sidereal hub since those were made long before we introduced the design change necessitating the new hub. Next photo shows the hub with the dial and drive disk blanks. Next is a photo of these parts mounted to the machine. Lastly the sidereal dial blank is shown with its circumference mirroring the adjacent plate cutout. 

Back Up Next