Back Up Next

Demonstration selector retrofit - finished, continue debugging  - March 2021

This month Buchanan continues the debugging process and continues to wrap up remaining items yet to be completed and finished. One project is the retrofitting of the demonstration selector. This allows the operator to select the demonstration of fifty nine complications including the entire celestial train where a twenty four hour period is shown for each turn of the demonstration crank, and a faster speed to demonstrate the orrery with an eye toward seeing the outer planets of Jupiter and Saturn's longer duration orbits, and then back to the normal operating condition where the clock drives the celestial complications in real time. Directional plaques are added to the planisphere's longitudinal grid.



These photos show the fabrication of the directional plaques that will be attached to the planisphere longitudinal grid lines: N, S, E and W.


Buchanan tried three different sizes to get the right proportions for the plaques in relation to the planisphere. Next the final plaques done in French silvering.

The plaques are positioned for trial fitting before they are silvered.

The next section deals with the partial retrofit and completion of the demonstration selector assembly. This will allow the operator to perform the demonstration selection with a key through the case glass via a hole; avoiding the need to open the case. All routine functions related to winding, setting of time, striking and celestial demonstration can be performed with the case closed using a crank or key. This minimizes dust infiltration. Optimally, the case will rarely be opened.

The original design called for the demonstration lever to be manually operated from the left side of the planisphere mask. But with the decision to have a case that would normally remain closed it was decided to eliminate that hand lever for a rotational key operated through the glass.

If one wants to reset any of the individual celestial objects, the planets on the orrery, set the planisphere, tellurion or demonstrate the prediction of a solar/lunar eclipse or adjust the other modules, the case must be open for the operator to manually move the appropriate objects or setting dials and levers.

The first step is to decide where the selector key will be located and how to integrate it into the existing selector plan where the control was to be located on the left side of the planisphere mask. Then to determine the location and shape of the selector dial. Shown here is a color photocopy of the clock. Given the placement of exiting components the best location was just below the red arrow and a sector dial shape would be obvious above.


The first photo shows the fabrication of a ball and socket click detent for the three positions available on the demonstration selector. A ball and socket provides a positive position for the gear selector to move within the transmission, (yellow arrow). Next Buchanan has to drill a retrofit hole for the updated selector assembly. At this point with the machine assembled and polished it becomes a very delicate procedure to drill into the frame; notice the bib below to keep the swarf away from the mechanism.


Here Buchanan performs another set of retrofit machining on a component of the demonstration drive. Again notice the effort to keep the part protected from the surface of the machine tool since it has already been through the finishing stage. Next updated selector components.

The first paper mockup of the demonstration selector dial is shown. We decided against the word "drive" for the default position of the selector, that being the clock in the normal position of driving all the complications in real time for the word "clock" which I had thought to be a bit more intuitive. The ideal ordering would have been the reverse, with the normative position being on the left, but the existing components made that impossible. There are many instances in horology where one does not wind the clock clockwise, but anticlockwise so this is not a meaningful issue.


The revised selector arbor unit is attached to the clock frame. The next photo shows the selector dial engraved right off the mill. Using CAD for dial engraving keeps all of the work accurate and uniform.


The selector sector dial done in a French silvering finish, next the dial hand is installed.


The entire selector lever work

Video of celestial demonstration lever.


These three photos show where a bug revealing itself as a tight spot in the selector pinion push lever. This lever operates within the grooved area shown by the white arrow in the first photo. A positive side thrust is provided to that lever by the U-shaped spring shown by the arrows in the next two photos. This works in conjunction with the three position detent to be sure that this pinion is always in the proper location for any of the three positions within the celestial drive transmission. At first the diagnosis was that the spring was providing too much side thrust, hence the tight spot.

It turns out that there was nothing wrong with the spring or design. There was some suspicion that this might be the case because before disassembly for the finishing process this bug was not there, so it may be a reassembly error or some alteration due to the finishing process. The latter is unlikely as these components are quite large in relation to whatever material would be removed in the finishing stage. One could see how this could be a factor in the realm of of a small, complex watch.

The answer was the flange of the sliding pinion was rubbing on the collet screws of the next gear over, indeed a reassembly issue. One can see Buchanan using a pointer to indicate where this was happening. Real estate within the machine, as one can see, is tighter than in New York City.

This bug and the photo above illustrates the immense importance of the debugging process over the entire machine. Look at how much disassembly had to occur to get at this one problem. The entire center section has been removed.

This is not unlike NASA launching people on rockets into space, if everything is not working according to plan, there is no turning back. It is unlikely someone else would be able to diagnose the problem and successfully reassemble the machine once delivered to its destination..

Back Up Next