Gear ratios

[Sub MandrelParticipant @submandrel]

This is more out of curiosity than intent to make a clock.

What are the usual gear ratios/tooth numbers used in a typical clock to give the 1:60 gearing?

I get the impression that, using cycloidal teeth, 48 and 60 tooth gears can be used with the same 12-tooth pinions, but I can't see how this would work.

What ratios would be used with involute gears to get the same result, or do you still use 12:60 and 12:40, but with different DPs?

Neil

[Sub MandrelParticipant @submandrel]

[Michael GilliganParticipant @michaelgilligan61133]

Neil,

Given that the question is "more out of curiousity" … you will probably enjoy plugging some numbers into this.

Thanks: You've given me a good place to park that very handy link for future reference.

MichaelG.

 

Here is another superb page.

Edited By Michael Gilligan on 06/11/2012 23:28:03

[Michael GilliganParticipant @michaelgilligan61133]

Neil,

Continuing to avoid directly answering your question …

It is worth noting that:

• Cycloidal Gearing works very well when Geared Up [i.e. when the output is faster than the input]; which is generally the case with mechanical clocks.
• Lantern Pinions; despite being generally classed as lower-grade, actually have a good cycloidal action.
• Most of what we think of as "Analogue" clocks are actually "Digital" in nature, because of the Escapement.
• Electrically driven clocks are an interesting exception; because the rotor, or pendulum, is actually driving the mechanism, not releasing it.
• The work done by Clifford, on magnetic drives and escapements, is fascinating. see here

MichaelG.

 

. added link to Clifford patents

Edited By Michael Gilligan on 07/11/2012 07:51:14

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Hi Neil ,

Lets see if a bit of engineering gets us anywhere :

(1) You can always devise a suitable gear train by writing down a randomly chosen set of gears which will work and then factoring , scaling , splitting and combining the initial ratios to get to where you want .

(2) There are ideal ratios for the sequential ratios in a gear train . Some of these are :

(a) The constant ratio series – all the same ratios at each stage .

(b) Incremental logarithmic and decremental logarithmic . Sequential ratios increase or decrease according to a logarithmic rule .

(c) Geometric series . Actually a form of (b) but more obvious and commonly used .

Each and many more have specific applications in different types of engineering .

(3) There are practical constraints on gear sizes .

(4) Heres an example of a geometric series using your 12 T pinion :

12/36 : 12/48 : 12/60   or inverse 3:4:5

Try all the methods in (1) to get a train to suit your actual needs . Note that within these methods you can change the number of gear pairs as well .

Regards ,

Michael Williams .

 

 

 

Edited By MICHAEL WILLIAMS on 07/11/2012 10:54:15

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Stub Mandrel on 06/11/2012 21:03:35:

This is more out of curiosity than intent to make a clock.

What are the usual gear ratios/tooth numbers used in a typical clock to give the 1:60 gearing?

I get the impression that, using cycloidal teeth, 48 and 60 tooth gears can be used with the same 12-tooth pinions, but I can't see how this would work.

What ratios would be used with involute gears to get the same result, or do you still use 12:60 and 12:40, but with different DPs?

Neil

Neil,

I had intended to set you off on your own voyage of discovery; but you did ask some specific questions.

Mr Williams has already responded to the Engineering points, so perhaps I should concentrate upon the horological ones:

What are the usual gear ratios/tooth numbers used in a typical clock to give the 1:60 gearing?

• There are many answers to this; largely dictated by the "quality" of the clock. Let me quote from J. Eric Haswell's 1947 book "Horology" … Discussing "Weight driven Regulator clocks" he says: A high numbered train is customary in the finest work, as follows: great wheel 192 teeth, centre 128, third 120, escape 30 and pinions of 16 throughout. Pinions of 12 are, however, sometimes used in which case the numbers become: great wheel 144 teeth, centre 96 and third 90.

I get the impression that, using cycloidal teeth, 48 and 60 tooth gears can be used with the same 12-tooth pinions, but I can't see how this would work.

• Yes, this is in the nature of the Cycloidal tooth form. If you want to delve into the geometry, I recommend reading "Gears for Small Mechanisms" by W. O. Davies. As the note on the back of the dust-jacket says: "Fine pitch gearing presents problems which are not solved adequately by scaling down copies of the gearing designed for power transmission. This book deals with those problems."

MichaelG.

[Sub MandrelParticipant @submandrel]

Thanks both Michaels,

At least I thinks so… my brane hertz now!

The atmos calculators look very useful.

I shall re-read the series on the month-going regulator clock and see if I can follow it proeprly now.

There's much of interest in this vein. I would like to make a 'grandfather' clock – the regulator adapted with a moon-phase dial, perhaps – but that's a manyana project.

Neil

[V8EngParticipant @v8eng]

Do not mention Grandfather clocks to clockmakers, I Was told many years ago (and many times) it's a Longcase!

[Michael GilliganParticipant @michaelgilligan61133]

Neil,

I think you will like this page

… the animation of a three tooth pinion is especially satisfying.

MichaelG.

[Sub MandrelParticipant @submandrel]

Hi V8

That's why I put grandfather in quotes!

Neil

[Sub MandrelParticipant @submandrel]

Michael –

The wacky gears further down are even more impressive!

Didn't Jacques Morel actually make something like that?

How on earth did you find that?

Neil

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Stub Mandrel on 08/11/2012 21:25:55:

Michael –

How on earth did you find that?

Neil

Can't honestly recall

When searching; sometimes it's a matter of "Use The Force"

MichaelG.

[Izack MaddParticipant @izackmadd89335]

Hi,

yes it it your resident utter nutter back for another weird and totally stupid question not really connected to engineering.

But this time it is connected. Clocks. More precisely moon phases. Hence the location.

No I've found the answer for once I just need someone to explain it to me. Some clever chaps, yes I'm English, over at IWC. What ever that means have realise they a "gear train"? I assume that's the tech name for a set of gears. Consisting of "30:90,32:90, approximates the synodic month as 29.53125 days. The 30-tooth wheel is on the day-of-the-week shaft which is incremented once a day at local midnight by a ratchet, thus turning at 1/7th of a turn per day. The second 90-tooth wheel is on the edge of the lunar display disk, and turns at 1/59.0625th of a turn per day, producing an error of 1 day in 122 years."

Now I'm about to start building a wooden clock so allowing for the fluctuations of humidity and temperature I'm sure that sort of ratio will just about do me.

But I don't grasp the relevance of the way it's written. "30:90,32:90". I realise the commas and colons indicate the type of connection just not what type. If it helps the original artic all is here "http://www.geartrains.com/gtwpgt.html".

What I need is a morons guide to what connects to what and how and does the diameter matter at all.

All help is very grateful and will aid my decline into total lunacy. Yes the joke is meant.

[John HaineParticipant @johnhaine32865]

See

http://www.csparks.com/watchmaking/CycloidalGears/RichardThoen.xhtml

For a fascinating article on why cycloidal gears should be consigned to history.

[Izack MaddParticipant @izackmadd89335]

Posted by John Haine on 23/11/2015 22:21:40:

See

http://www.csparks.com/watchmaking/CycloidalGears/RichardThoen.xhtml

For a fascinating article on why cycloidal gears should be consigned to history.

Hi,

Thanks for the great link and fascinating info. You don't happen to have a link with a image of the different tooth design at all. Only as a dyslexia advantagee I find I learn at supersonic speeds visually rather than the more Neanderthal speeds I attain reading and when combined the learning curve goes into light speed.

And for those daft enough to care that means by my hypothesis that I'd be thinking faster than time so would have sorted the answer before I'd seen the image but in order to maintain temporal balance I still need to see what shape the teach should or should not be.

Now you understand why I realistically label my self as an utter nutter.😜

[Izack MaddParticipant @izackmadd89335]

Hi,

Again, silly thought but on a mechanical clock powered by weight alone. If it had a moon phase indicator would that have to be disengaged to adjust the clock for daylight savings time, and back again. Or should the phase alter as well.

You'll have to excuse my slow wittedness only the orderlies have just been round force feeding us the Wappy and NightNight pills. 💤

[BazyleParticipant @bazyle]

Izack, The target is 29.5xxxx days for the moon to wax and wane but to make it look better it is sometimes done by having two moon pictures on the circle and move half as fast, ie 59.xxx days. The gear train is 1/7 provided by the cunning use of the day of the week reduction followed by 30/90 then 32/90 ie (1x30x30)/(7*90*90). The last 90 gear is on the moon picture wheel.

Size doesn't matter provided you can fit it all in.

For changes to daylight saving a traditional friction drive on the minute hand allows the hour to be moved without much affect on the going train or the mechanism above because the date advance ratchet isolates it. Chiming and alarm mechanisms are more tightly linked to the going train which is why you might be worried about a problem.

Are you going to use the Brocot perpetual calendar?

[Izack MaddParticipant @izackmadd89335]

Posted by Bazyle on 24/11/2015 01:42:49:

Izack, The target is 29.5xxxx days for the moon to wax and wane but to make it look better it is sometimes done by having two moon pictures on the circle and move half as fast, ie 59.xxx days. The gear train is 1/7 provided by the cunning use of the day of the week reduction followed by 30/90 then 32/90 ie (1x30x30)/(7*90*90). The last 90 gear is on the moon picture wheel.

Size doesn't matter provided you can fit it all in.

For changes to daylight saving a traditional friction drive on the minute hand allows the hour to be moved without much affect on the going train or the mechanism above because the date advance ratchet isolates it. Chiming and alarm mechanisms are more tightly linked to the going train which is why you might be worried about a problem.

Are you going to use the Brocot perpetual calendar?

I'm still in the planning staget at the moment. The plans I have are for a basic clock with no moon or date mechanism and as such will require a separate weight. This means I have a dedicated powers source for this side of things taking the syncing from the hour had and a 14 tooth gear. Which will be tripped every twelve hours so giving me the needed 1/7th of a turn. As for the Brocot perpetual calendar I'd never heard of it so thanks as I was just going to fidure it out my self but if someone else has done it that will be easier to adapt. As I was puzzling over how to sync the advancement of the date wheel for February and a leap year. So I'll be googling Brocot. I wasn't thinking of adding chimes so the only issue regarding DST was a foggy thought. But thanks for setting my mind at rest. A great help

[Izack MaddParticipant @izackmadd89335]

On the subject of clock making and morons. Which covers me.

If I use Lignum Vitae for either the shafts or arbour.

Which is the best shaft or arbour?

And what's the best material to use for the other half? Or should I use LV for both?

[BazyleParticipant @bazyle]

Even in the most desperate times I think impoverished wooden clockmakers (and watchmakers in Russia) still used steel for the actual bearing part of the shaft but used a self lubricating wood like LV as a bush in the plate, or as used by Harrison as small wheels on steel pins to build a low friction lantern pinion. So you are looking for a straight grained hardwood that won't warp and drill axially to insert a steel pin each end. While oak springs to mind also consider that it promotes rust owing to the tannin content (which is what also stops it rotting). You might investigate teak. Wood against wood is no good.

For the steel clockmakers use a version known as 'pivot steel' which is a high carbon hard steel but amateurs tend to just use silver steel. Owing to the moisture present in wood rust is a problem and stainless steel would be an advantage. You might look at darning needles and the like.

[Ian S CParticipant @iansc]

My only refferenceto wooden clocks is one published by "Popular Mechanics", their suggestion is to use quarter sawn hard wood, Black Cherry is the favoured wood. Quarter sawn wood swells and shrinks the least in width. I suspect that the steel pins used in the end of the wooden shafts are nails, the bearings are short brass tubes pressed into the wooden plates. The gears are cut by hand.

Ian S C

[Author]

Posts