Jump to content

[TC28] The Marvellous Mechanical Date Engine (was Magic Calendar)

aeh5040
 Share

Recommended Posts

aeh5040

aeh5040

Posted
  • Author
Posted
1 hour ago, SAM1626 said:

just wondering how many pieces is this currently and where did you get the special colored shock absorbers that are in this picture?

Pieces: no idea - maybe 2000?

Gold springs https://www.bricklink.com/v2/catalog/catalogitem.page?id=236062&idColor=11&ccName=6472027#T=C&C=11

I'm quite pleased how the colour scheme is turning out. At first I was disappointed that the most neutral colour for the 14t splat gears is medium nougat, but then I realised it is quite a good steampunk colour, especially combined with some pearl gold.

aeh5040

aeh5040

Posted
  • Author
Posted (edited)
2 hours ago, Davidz90 said:

Looks great! Love the color scheme.

Thanks! Pity the trans blue is all going to get covered!

8 hours ago, SAM1626 said:

Not going to pretend I could do this but this is very cool and I can't wait for the final results

Many thanks for the encouragement!

Edited by aeh5040
aeh5040

aeh5040

Posted
  • Author
Posted (edited)

Right. Everything is now assembled. Not completely surprisingly there was still a bit of a problem with gear lash, meaning that the output day of week indicator would not be reliably centered on the day. Acceptable, but not ideal.

I have therefore added a further detent mechanism for the output drum. This one is a much less aggressive gravity driven one, compared with the strong spring driven ones for the inputs. Astonishingly the gear ratios work perfectly for this, and there was just exactly enough space to fit it in, one of many examples of tight tolerances in this model. It turns out to work really well!

One drawback is that if you want to calculate the day of the week on the space station, it may not be so reliable! :laugh:

800x601.jpg

Remaining task is the stickers...

Edited by aeh5040
aeh5040

aeh5040

Posted
  • Author
Posted (edited)

Working on a taller window for the day drum to display the result more prominently.

Some tolerances are incredibly tight, and it is really lucky that they work out. The drum cannot fit beside the middle hole of the thin red left arm - there is only enough room at the two ends. And the clearance between the drum and the grey (soon to be red) engine crank part is so small that the paper sticker cannot go on this bit! Also the liftarm sandwich at the bottom needs a gap to accommodate the crank part.

800x601.jpg

Edited by aeh5040
aeh5040

aeh5040

Posted
  • Author
Posted
3 hours ago, 1963maniac said:

It's looking good! IWaiting for it to be revealed what it is in it's entirety.

Thanks! Sorry for all the teasers! Getting the stickers on correctly is a delicate business that requires a fair bit of disassembly...

aeh5040

aeh5040

Posted
  • Author
Posted (edited)
8 hours ago, ord said:

Robust looking switching - it seems to lock into positions nicely. Can't wait to see the full video.

Thanks everyone! Indeed, you have exactly identified the central challenge here. Due in part to quite a lot of good luck, I think I have managed to solve it...

Edited by aeh5040
  • aeh5040 changed the title to [TC28] The Marvellous Mechanical Date Engine (was Magic Calendar)
aeh5040

aeh5040

Posted
  • Author
Posted (edited)

A few more details:

The algorithm is basically like the tabular methods at https://en.wikipedia.org/wiki/Determination_of_the_day_of_the_week and physical "perpetual calendars" like this one https://en.wikipedia.org/wiki/Perpetual_calendar#/media/File:50yearcalendar.JPG.

However, so far as I am aware, all previous such methods and devices required some sort of manual adjustment for leap years.  This device handles leap years automatically.  As you can see, the year drum has 14 positions, alternating between leap years and non-leap years, and its motion controls the selector which uncovers different instances of Jan and Feb (other months are not affected). 

A second innovation is the way exact centuries (1900, 2000, ...) are handled.  There is no "00" on the year drum.  Instead, these years are formed using the "0" and "000".  These are positioned so as to get it right (leap year for a multiple of 400, but not for other multiples of 100).  More centuries beyond 1601-2300 could be handled by adding extra entries to the century dial, but I decided this was enough.

The day, century and month drums also have 14 positions rather than 7.  There is no fundamental need for this, but it simplifies the construction and allows for fewer entries sharing a position (in particular no sharing on the century drum).  The output day drum moves by a full 1/7 turn, partly to make it more dramatic and partly to combat gear lash issues.

It would be possible to add manual horizontal sliders to the drums so that a single entry could be selected, making it more like a calendar.  There is no obstacle to this but I decided it would make it too fussy.

Regarding mechanics, it is important that the signals are constantly geared UP, to minimize the effects of gear lash.  Gear lash is always a constant fraction of a tooth for every gear meshing (even more for a differential), so each "click" needs to correspond to a big rotation of the shafts, so that the lash remains a small fraction of it. In fact each click corresponds to one rotation for most of the mechanism.  Of course there is a balancing act, because gearing up too much would increase resistance. It is also important to minimize the gears.  All this ensures that the lash in the final output is fairly small even without the final gravity detent, so then the gravity detent is enough to click it into place.

The splat gears are very useful both for their large size (so that they can mesh with an axle that does not intersect the drum) and to get the necessary factor-of-7 gear ratios.

The "stickers" are actually not stuck to the drums, but form a tight band around them held in place by friction.  This makes adjustment much easier.

There are lots of tight clearances everywhere - it feels rather miraculous that everything fits.  The drums themselves are quite close to each other; the vertical axle for the leap year selector passes really close to the drums; the taller red window for the day involves some ridiculously small gaps as described above.

I'm really happy with how it has turned out. Many of us know that with a model like this there is almost always something that never quite works as well as it should. For once I seem to have escaped that curse!

Edited by aeh5040
aeh5040

aeh5040

Posted
  • Author
Posted (edited)
5 hours ago, Davidz90 said:

Absolutely beautiful! The concept, looks, delightful sound of the cycling detents...

Thanks! Glad I am not the only one to appreciate a satisfying clunk :wink:

Edited by aeh5040
  • 2 weeks later...
aeh5040

aeh5040

Posted
  • Author
Posted (edited)

Not having anything else to do, I made a digital model, and instructions are in preparation...  Also some better quality pictures below.

800x749.png

800x700.png

800x544.png

800x649.png

800x713.jpg

800x770.jpg

Incidentally, the "perfect fit" geometry was used in various ways to position the red windows:

800x906.png

Edited by aeh5040

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
 Share
Go to topic listing

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...