Back Up Next

Orrery, continue inner planets gearbox  - March 2019



Buchanan writes: I have the first arm clutch complete. It seems to be working well so I will proceed with the other three. It took a little squeezing to fit in but is better than doing it outboard on the outer arbours. The outboard method has the danger that other planets could move when adjusting one. The clutch is basically a friction design as represented by the wavy steel ring in the center of the third photo. It is a design that has been used dozens of times in this project.


The other three arms are now prepared for Mercury, Venus, Earth, and Mars.

Buchanan writes: Here is the process of making a very thin spring washer out of steel shim stock 8 thou thick (six photos below).


First it is sandwiched between two brass plates and drilled undersize, (first photo). The inside diameter is bored out to the correct diameter using an adjustable boring head, (second photo).


The washer is cut out with snips to roughly oversize and placed on a stepped mandrill on the lathe. The upper piece of the earlier brass sandwich is used to clamp it against the mandrill, (third photo). The tailstock live centre is used to provide clamping force to drive the washer, (fourth photo).


The washer and everything else is turned to the correct outside diameter, (fifth photo). The completed washer and a few interesting leftovers, (sixth photo). The clutch washer is made wavy with the careful use of a pair of pliers.


The four clutches are now complete.

At this point it became apparent that the design by Philip Hahn did not provide for the Earth to revolve. I had resigned myself to this fact since the orrery was being built to the specifications provided by documents left by him.  

Buchanan wrote: I have given the Earth rotation some thought. I have done a feasibility study and recon I may just be able to take a drive from the Earth’s moon arm which does one rev every 27.3 days and step it up by the correct ratio of 27 point something to get our world spinning with a tiny 3 pair gear train, mounted below the moon arm, on the main earth arm. The Earth is so small and weighs so little I think this is possible. I have checked on the max tooth count possible with my smallest cutter in the space available and I could fit a 90 tooth wheel in. It seems with a minimum wheel size of 10 teeth and a max size of 90 teeth and 6 gears we could get an accuracy of about 1 day in 50 million days. It will be a bit like a cherry upon the cherry on top of the cake. We have about an inch disk of space that I can use so it will be pocket watch gearing. Once again Buchannan has shown ingenuity to solve a problem. In this case a pretty difficult one since the solution must be worked into the existing design and parts already created with little room to maneuver.  



Buchanan writes: I have the Mars eccentric bearing and jewels ready and mounted and the Mars gear spoked. Next are the arbour and the offset for the planet rod. I will complete Mars, Venus (simple) and Mercury so that I can see what space is left for the Earth gearing.

The calculations are a little funny for the earth gearing as I have to work it out from the difference in angular velocities between the moon drive gear and the Earth arm. Both which are rotating around the sun. Then I also have to take into account what direction everything is turning. And finally arrive at an earth that spins once a day in the correct direction.


The Mars armature is now seen mounted to the inner planets gearbox.

The last set of dial artwork is ready to deliver to the Chinese enamellers. Each has the paper design as well as that design on a plastic ring that has been machined to the curvature needed for the copper blank. Clockwise from upper left is the corrected month/season dial, redone demo/international time dial and the redesigned orrery dial ring. The fact that these are the last dials to be done is another indication that the Astro-Skeleton Clock project is drawing to a close.

Buchanan writes: I have the bridge practically complete. It carries the sun arbour and is attached to the bottom frame; it provides a central support to the centre tubes as well. There is the cosmetic finishing to do. This is a steel bridge just for contrast. Buchanan goes the extra effort to make the bridge from steel for visual contrast; making the part from steel is more difficult to fabricate and finish than from softer brass stock, (see photos below).


The steel bridge has been milled to shape from the stainless steel blank and is being cut from the blank with the slitting saw, first photo. Next the bridge is separated from the blank.


The first photo shows the rough bridge checked for fit on the lower inner planets gearbox frame. Next the bridge is within the frame, yellow arrow.


 The first photo shows the fitting for the sun arbor and the next photo with that arbor mounted to the bridge.


With the center arbor fitted, the concentric tube assembly can be supported and in turn the four planet armatures fitted. As has been done countless times in the past the ungainly, flat rectangular blanks will later be machined into slender, decorative parts. Buchanan writes: I will now continue with the Venus and Mercury arms before I finally return to the Earth arm. I want to maximise the visual appearance of the Earth gearing and need to know exactly where the surrounding planets are.


The artwork for the four planet arms is shown in the first illustration. Next the beginnings of the mercury armature.

Buchanan writes: I cut the centre gear for Mercury (second photo). I increased the length of the arm slightly and increased the tooth count from 57 to 72, another 15 teeth. About a 0.15% increase in overall tooth count! I hope to have Mercury arm complete and Mars gearing done next. Venus is a simple arm, no gearing; its orbit is only about 1% out of round, far less than the Moon.


The staking tool is used to push a large jewel into the center wheel on the armature for Mercury. In the next photo one can just see it below the screw on the center wheel. The entire armature is about 1.5” (4 cm).


Three of four armatures are complete, from he top: Mercury, Venus and Mars. An exploded view of the Mars armature is shown in the second photo. Earth is next to be made and presents more complexity due to the Moon as well as the decision to make the Earth revolve.

All four planet armatures are machined to their curvilinear design and displayed within the context of the entire orrery, the upper center planets gearbox frame is removed. 

This photo shows the entire compliment of the orrery drive wheels along with the two completed planets of Jupiter and Saturn. One can see how this module has nicely filled out the center space between the two remontoire fly fan towers.

One can more easily see the fit of the orrery within the rest of the machine. The entire bulk of the module is within the vertical elevation of the two upper dial sets, but just below the superstructure of the two remontoire towers. The height also allows one to see the operation of the upper section of the compound pendulums as well as their beat plates; this is important since the lower section of the pendulums is largely obscured by the base of the clock. The orrery is close enough to the top of the machine to make it look like a continuous extension of the wheel works. If the orrery were to be raised by even an inch or so it would have looked ungainly; like it was mounted on a stalk balanced over the rest of the clock.

This is serendipitous since the superstructures of the remontoire towers would have presented a serious obstruction to Saturn’s orbit. Making the entire orrery smaller could have been done, but we are working in some areas at the scale of a pocket watch, the limit of Buchanan's tooling. Furthermore one begins to lose detail of the components when they are too tiny.

One of the earliest decisions made was that of the overall dimensions. In early 2006 there was some discussion about making it about one third smaller. We were fortunate not to have chosen this as it is likely that the smaller wheels in many of the complications could not have been made. But more important we would have lost some of the ‘presence’ that the machine commands from the fact that it presents itself to the viewer with detail that is easy to discern without magnification. It reaches out to the audience.

Here the orrery melts into the rest of the machine. Of course the dial ring will bring definition to the space occupied by this module.


This video shows a demo of the entire clock with a focus on the orrery as of April 2019. In this view only the outer planets gearbox is functioning.

Buchanan now turns to the Earth armature to improve on Hahn's design and make it revolve in both operational and demo modes.

The drawing above illustrates the new components Buchanan will create and where they will be located to make the Earth rotate. All of the new components are attached to the original Earth armature colored in yellow and extending from the central stalk. Buchanan explains below.


I have worked out a set of 6 Earth gears that I can fit into the space available. The accuracy of the Earth’s revolutions will be 1 minute in 186 years, or an hour, 1/24th of an earth revolution in 11,113 years. The green parts are the new gears, the  red, jewels. The yellow is brass and blue is steel. I will rotate the gears 90 degrees in relation the earth arm so that they clear Mars. I just can’t draw in 3d yet. I have also done a check on the orrery's general direction of rotation of planets and moons and all is correct, and matches the Tellurian. I will now start to cut the last 8 gears for this clock. The last 8 wheels! I can hardly believe it after all this time.

A comment on my calculations.

I have used the values given in the book on Hahn for the input speeds on the earth gearing. Hahn has inaccuracies in his moon orbit speed as well as his length of year. I have calculated the earth speed compensating for these discrepancies.

Hahn’s earth arm is I rev in 365.242 794 Days

Correct period                       365. 256 363 Days

Error 19 .5 minutes a year

Hahn sidereal moon period is 27.321 589 days

Correct period                            27. 321 661 days

Error 1.85 minutes a year

Our earth error 1 minute in 186 years.

I expected Hahn’s calculations to be a bit off, but am a bit surprised by the error of 19.5 minutes per year. The 1.85 minutes I could understand given the time when he worked. I do like 1 minute in 186 years better! It will be interesting to see the motions of the Earth, moon and inner planets of the tellurian move in concert to their counterparts in the orrery. Greater gear and pinion counts than Hahn used also contribute to greater accuracies.


The first photo shows the meticulous hand alteration that Buchanan makes to a standard jewel. The standard jewel is mounted to a mandrel and is ground down on a copper rotating disk charged with diamond paste. Nearly all of the jewel’s diameter has been ground away changing the shape from a doughnut to a slim sleeve. These two sleeves are used in the upper and lower sections of the cannon pinion shown in the second photo, (the moon sleeve).


Buchanan writes: This is the moon sleeve and its drive gear. It also carries the 91 tooth gear that still has to be spoked. The hole in the centre of the steel part is for the earth arbour. I will also fit an end stone for the earth to stand on.




Buchanan writes: I am painting doors faster than floors some days.

I had a pinion practically complete (first photo), with steel pivots added, and was machining the last pivot to size when the tool caught and you can see the result, second and third photo. This pinion is just under 1/8 inch diameter and 16 teeth.

So I painted a door. There was just enough not damaged to be able to use it in a new position, fourth photo. I moved it from one of the outboard arbours to the earth arbour, fifth photo. No paint on my feet, but, mud on my face. 

I have also spoked the two "large" 91 tooth wheels today, 5/8 inch diameter .2 module, last photo.

I reply: I think that damaged pinion would have given the clock that “antique look”.


The first photo shows a steel pivot being cut just before parting from the steel rod blank. Next a second wheel which will mate to that wheel is being prepared for cutting out of spokes. The jeweler’s saw blade passes through a 0.7mm hole to begin the cut.


Once the raw cut is made for each sector,  Buchanan finishes up by hand with a micro-file. Next the same wheel is shown to the left. Its mate, to the right is the same wheel shown in the second photo above.


The staking tool is used to press-fit the two tiny wheels together with the result shown in the second photo. The exacting attention to detail results in wheels less than 1/8”, (0.5cm) having spokes. I wonder how many other clockmakers would have done this. These are amongst the smallest wheels in the clock.


The wheel pair is shown in the lower right sector with other parts that will comprise the Earth gearing, first photo. Next a depthing tool is used to check for proper meshing of the wheels.

Buchanan writes: This is the layout for the frames for the Earth gearing.  I have done the test depthing of the wheels so I have the correct dimensions for setting out the pivots.

There has been a lot of juggling to obtain a practical set of frames that can be made and also fit into  the standard arm design so that it does not look like an add on. The process will become more apparent as I continue, it is hard to visualize in my mind, let alone explain on paper.


These two photos show the counter-sinking and mounting of the steel foot bush. There won’t be much of the flat foot left in the end.


Buchanan writes: I have got past the design bottleneck. I thought everything was all right with my design but forgot to take into account the large middle idler wheel that brings the drive out from the centre. So I had to go back and start again. The sequence of assembly has to be taken into account as well as the 'pretty’s' and mechanical requirements.

I will send a final scan of the frame design. I have a main arm frame the steel foot bush and a lower extension cock and finally a small bridge on two pillars across the outer two gears. I think it will look good. Photo 605, the last photo, gives a general arrangement of wheels. I also want to include an age of the moon dial the same way as the other planets have an eccentric orbit dial. I have a mounting method worked into this as well. The design illustration when compared with the one first created for the Earth armature extends the curvature past the second ivy spur detail to include five additional jewel pivots. This is the final design drawing for the clock movement!

I am now working on the extension cock. There are different heights of bearing and pillar pads to take into account, but I think the end result will justified.


Buchanan begins the fabrication of the frame work for the additional components necessary to make the Earth revolve. In the first photo the steel foot for the Earth is prepared. Next the mount is cut to fit the curvilinear contour of the armature.


In the first photo one can see the scribed outline of the ivy design for the armature. The two pins on either side of the center tube will eventually be cut down to serve as location pins for the frame that will hold the new components needed to make the Earth revolve. Next an additional brass plate with several pre-drilled holes is positioned next to the armature. The locating pins now run through this plate ensuring exact positioning of the new plate to the armature. Along with the steel foot, all of the new components are located on this plate.


These two photos show various machining operations. They concern the fabrication of a cock for one wheel set. Look carefully at the tip of the cutting tool in the second photo. It is at the base of the nascent cock; note that it is fabricated from the thick, solid plate piece and will not be a separately screwed in part as would be in a standard design. There is not enough room both laterally and height-wise to accommodate the cock’s foot and associated securing screw. This requires additional machining and reflects the innovative way Buchanan solves a problem that was created because there was an ad-hoc change in the design and function of this component.


Further machining in done to make the attached plate match perfectly with the Earth / Moon armature, first photo. Next the new wheels are checked for fit.


In the first photo the plate is cut to fit and extend the contour of the armature. The second photo shows the new component frame and is all that is left of the original, thick plate.


These two photos show the new frame component and its compliment of bearing jewels. The three empty holes seen in the second photo represent the two locating pins with the center hole for the screw that will secure this frame to the Earth armature. The entire piece will be substantially thinned as it is finished by hand to its final shape.


The first photo shows the reverse side of the completed and mated frame and armature. The second photo shows just how small all of this work is. The tiny wheel pair seen earlier is located just to the left of the lower wheel’s pinion and is on the scale of watch work. There is still an upper bridge needed to complete the assembly.

The orrery center planets gearbox as of March.

The Earth and the new components to make it revolve within the context of the orrery, inside yellow circle.

The machine as of the end of March 2019. If one thinks there is room to spare within the context of where the orrery is positioned, look carefully on the left where Saturn is positioned. 

Back Up Next