Antikythera Mechanism

[Mechanism Man]

Hi Folks,
I've just found your forum and don't really know how I've missed it for so long. I've been messing about with wooden gears for the last decade or so and probably could have learned a lot by having a browse here once in a while.
Anyhow, this might be a bit left field compared to the rest of the posts here, but I've finally, after several years of planning, research and generally putting it off, gotten around to designing and building a wood geared version of the Antikythera Mechanism. Not strictly speaking a clock, but it most certainly can be used to measure the passage of time - just over a longer period than a normal clock...
For those not familiar with it, it's the 2100 year old gear driven shoe box sized machine that was discovered in very bad condition within a shipwreck off Greece in 1900. Since that time researchers have discovered that it could predict planetary movements (including retrograde motion where the planets appear to move backwards in the sky), eclipses, passage of time, moon phases, and all with an accuracy that in some cases has less than a 1° error over a period of 500 years... and it pre-dates the first gear driven clock by over 1000 years. In other words, it's pretty darned clever.
Latest research has put the gear count inside the machine at something in the region of sixty, and it would have predicted the motions of all the planets out to Saturn with a really neat and elegant system of gear trains.
It is also more or less unique in that it represents what can be seen looking out from Earth - so Earth itself is not featured on the machine - but in a lot of ways it makes it a much more useful device than a standard orrery, and way more complicated...
So, my plan is to build a version of this incredible machine and incorporate all the latest thinking on the gear trains which nobody has yet done in a physical machine, all the machines so far built are modeled on earlier thinking. And I'm going to do it all with hand cut wooden gearing - not a laser or CNC machine in sight...
The machine will end up about one and a half times bigger than the original, but it's still quite small considering what it will eventually do.
What could possibly go wrong?!
I've already made a really good start and got pretty much all the gearing cut and trial fitted to the main set of frames. It doesn't yet run, although all the gear pairs have been test run successfully within the frames, the assembly in the photo was just to see if it all fitted together as it should - and it did! What you see represents about a year of work I guess...
There is still lots to do, but I think that I'm definately heading in the right direction.
I'll try and keep you all posted as I progress.

[millerdlca]

That is .... stunning, especially with hand cut gears.  Looking forward to seeing more!

Dennis

[steve323]

That is a really incredible looking machine.  I hope you continue to post progress updates, even with the light traffic on this forum.

Are the results shown by the positions of the 3 clack rings?  Or will they eventually end up as hands coming through the brass tubes on the front face?

Is the fastest rotation time equal to a single earth day, so that the positions of the known planets and moon would all be displayed on the front face?

thanks
Steve

[Mechanism Man]

Hi folks, thanks for the encouraging comments. It's nice to know that there are other shed bound people like me out there who appreciate a gear or two!
Steve - The display outputs from this are many and various. I still have to make up the graphics for them, but this is what I have to achieve:-
On the front (the back of the machine in the photos) there will be a full display that will show the exact position in the sky relative to the constellations of all the planets out to Saturn; the moon (and it's phases); the suns position; and finally the date that the display refers to. The Earth itself isn't featured because the machine shows what can be seen, and where in the sky to look for it, whilst looking out from Earth, so it's more useful than a standard orrery.
On the back - the side which is visible in the photos, there will be two large spiral displays that will be divided into months (235 months for one of them) and will help show when we can expect to see eclipses of the sun or moon, plus three other lesser dials that show other measures of time and adjustment values for the time of day to expect the eclipse - it really is very clever considering it's 2100 years old... In total there are 13 different outputs - or 14 if you count the Moons phase.
if you want to read more have a look at the Antikythera Mechanism on Wikipedia - their page on the machine is brilliant.
Below is a photo of around three quarters of the gears freshly cut - that really was a job and a half...
I'll post more photos as it progresses.
:-)

[steve323]

Thanks for the update MM,

It is amazing what they were capable of building 2100 years ago.  Even more amazing is that they were able to plan out a gear train with that much complexity just based on observations of the sky.  And all of this was done before telescopes were invented.

Steve

[Mechanism Man]

Sorry for the almost total lack of progress on this in recent weeks, other projects and life keep getting in the way...
in the mean time I have also taken on building a nine planet wooden orrery containing 49 gears - to be finished by September 21st... No pressure there then! Still not a clock I'm afraid, but it does keep time in a way I guess, so hopefully it's not too far off the beaten track.
So I shall change the direction of the thread for a few weeks and chart the progress of this instead - I'll go back to the Antikythera Mechanism once this is done.
I need more hours in the day....

[Mechanism Man]

Spent the last few evenings wading through the gears for the new orrery comission. Getting there, but it's slow going. There are 47 in total I think.
Sooo many teeth...

[Sablatnic]

The more I see the more impressed I get - I've hand cut a few gears, bur this is most impressive!!!

[Mechanism Man]

Almost there - another couple of hours should see the last few 19 tooth gears cut. Oh yes. :-)
I then have to go back through them and cut all the spokes, and sand them, and apply varnish... Doh...

[Mechanism Man]

Hooray! All the hard work is now done - all the gearing is now cut, with the last few spoke holes finished about 10 minutes ago. All that's left is sanding and varnishing to finish them off and then it's time to cut the frames.
The gear cutting is by far and away the longest job, but if you take a bit of time with it and get it right then the rest of the machine goes together so much more easily.
More soon.
:-)

[jltrent]

Impressive start on the Antikythera Mechanism!  More so in that the gears are handcut!  I was "bit by the Antikythera bug" and attempted a build a few years back (see thread further down in the "General Discussion" section), and was about to pick it back up again a few months back when I noticed that new research had come out from Tony Freeth on the superior planets, involving further sets of epicyclic gearing.  This caused me to pause, so I began to gather supplies, in particular semi-transparent plexiglass for the gearing and the 9 concentric tubes for the planetary and lunar outputs, but other projects and life events got in the way...

On my original model, I found the d1/d2 gear pair to be problematic, as in one direction the gearing up would torsion the arbor causing jamming.  (Interestingly enough, in Derek de Solla Price's diagram, he shows a spring on the d1/d2 arbor, which probably is the compensating mechanism for the torsion...) 

It's still hard to believe that the design of the Antikythera Mechanism existed 2100 years ago, both even more so, that the techniques for building it existed.  I can easily see the design evolving from the ancient mathematical methods of geometry, but how they went about building 9 layers of concentric tubing for the lunar and planetary dial seems to me to be a much greater technical challenge...

In any event, am interested to see how your rendition eventually comes out once you again find time, and to hear some of the challenges you had to overcome...

R, Jon

[Mechanism Man]

Thanks Jon,
Yes, for a 2000 year old design it really is quite something. I understand completely what you say about D1/D2 - potential for twisting here is quite high, so although it doesn't look like I'm going to experience the same issue, I shall beef it up a bit, just to be sure.
And as for the nine layers of nested tubing, I too think that this would be the biggest problem that they would have to to have overcome with the original - maybe it was done with separate pointers as Professor Evans' team have suggested? But that's too messy for me, and I have access to nested tubes! I can't wait to get the orrery finished so that I can get back into this - another month and I'll be there.  :-)
Anyhow, here's a photo of all the orrery gearing finished and ready to be fitted - that was a long month... Frames next.

[Mechanism Man]

I've got gears, and as of yesterday I've got a full set of frames. Must be almost time for a bit of assembly!

[jltrent]

Orrery is coming along nicely.  Great progress considering it's all done by hand!

Regarding Prof Evans theory on the separate planetary pointers, I seem to recall his suggestion was only topical (ie, it didn't offer proposed gearing) and came out before Dr Freeth's proposal that made sense of the four (4) spoked b1 gear.  Speaking of b1, I took a look again at the fragment A radiograph, and noticed that the concentric "b" arbors appeared to include a square tube.  Am wondering if alternating square and circular tubes might have simplified the engineering, and simultaneously reduced the friction to only four (4) touch points between each tube?!

I also recall when building my earlier model my initial confusion of the pin between the "e" arbors that produce the elliptic of the moon phase.  Only after building the epicyclic gearing did it become clear that the two (2) "e" arbors only wobble a few degrees from each other, and I supposed the pin must have been some type of safety, to reduce the chance of significant slippage between the "e" arbors, or to aid in taking pressure off the epicyclic gear train...

Will be interested in seeing your orrery once complete.  You're moving along smartly...

R, Jon

[Mechanism Man]

Hi Jon, assembly has now started in earnest and is going really well so far. Just hope that the rest of the build goes as smoothly - the 20th September deadline is getting really close now...  :-/
I can hear the Antikythera Mechanism calling...