Basic Clock Design

.

Agreed ... It's one of several favourites in my collection.

MichaelG.

I find many books on clock making disappointing. The one with 2 designs in it I mentioned sums it up really - aimed at clock makers. He tried to avoid that but I'm not convinced he succeeded.

My collection is scant. Practical Escapements and Wheel and Pinion Cutting plus one modern book on Watchmaking. Plus djvu's of a few old ones of various dates. The one called The Modern Clock isn't bad actually it's mostly written pretty clearly and covers and or mentions a number of aspects. That one dates from around 1900.

On Harrison the most interesting thing I have found of late is the series of video's I linked to and another of one of his actual wooden clocks.

PS My Vienna has a lion no signs of eagles. Not sure what should have been on the other one

John

-

Edited By Ajohnw on 18/04/2016 23:20:18

Edited By Ajohnw on 18/04/2016 23:21:41

Edited By Ajohnw on 18/04/2016 23:22:51

If you hunt about on the netty you will probably find a load of out of copyright horology stuff

Personally I feel that there is too much mystique associated with clock making. Read a book and in many respects all they give is some ones opinions and in some cases red herrings of a higher mathematical type. The word type is oh so close to hype.

Beginners clocks are an interesting subject that I feel is summed up pretty accurately in the book with 2 designs in it I mentioned. He states that the 2nd clock is more complicated but takes exactly the same skills as the first one but as there is more it will take longer to make. In real terms the same applies to the Burgess clock.

The old books can have their place as they can contain techniques and information that isn't available anywhere else. As a for instance I noticed 2 methods of cutting escape wheels and ratchets. Modern a form cutter is used but it's also possible to use 3 cuts with slitting saws - old.

All of that is some what exaggerated in places to make a point.

John

-

Thank you for all of the comments and suggestions to date. As yet, I am still undecided about a remontoire, but shall study them in more depth before deciding. Although it seems the simplest, I am not keen on an electric device, simply because I do not wish for the clock to be reliant upon batteries or to need plugging into the mains.

I have revised the design, using 0.8 mod as a compromise size between 1 and 0.6 mods. With 0.8 mod, the pinions are large enough for me to have a fair chance of constructing them as lantern pinions. Anything smaller and I doubt that I would have much luck.

I have decided to use weights in the end, not spring.

The attached drawings show the general layout, but the frames are simply representative, but finalised, as are the long arbours at the front.

I plan to use a one metre pendulum with a thirty tooth escape wheel. The tooth count and ratios that I have settled on, starting with the escape pinion are:

10: 75

10:80

10:30

8:32 (the hour wheel)

The wheel on the winding barrel is 120 tooth

Do these ratios and general layout look reasonable?

Next step - pendulum and suspension. Can anyone direct me to any links or sites which cover this area? I can work it out by studying other designs, but some advice is always a help.

Regards,

James.

Edited By James Alford on 20/04/2016 07:31:02

Edited By James Alford on 20/04/2016 07:31:54

Edited By James Alford on 20/04/2016 07:32:49

Edited By James Alford on 20/04/2016 07:57:45

Better to move the winding barrel to one side to keep it away from the pendulum. Your pallets embrace 6½ teeth which is rather short. 10½ would probably be better.

Lots of different designs for the pendulum/crutch assembly. The usual suspension is a short strip of thin spring steel. Ready made suspensions are available.

I think the easiest escapement to make and adjust is the Brocot type.

Russell.

Edited By Russell Eberhardt on 20/04/2016 09:22:50

Well if you are using weights as your power source you don't need a remontoire.

regards Martin

But, unless you are prepared to stop the clock to wind it you will need some form of maintaining work.

Russell.

.

Martin,

Just for clarification ... When I originally suggested that James might consider using an electrical remontoire, it was in the [widely accepted] sense of 'electrical rewinding', and was specifically to allow the use of a small driving weight on his intentionally delicate clock plates.

There is inevitably some 'blurring of the edges' when it comes to defining what constitutes a remontoire ... which is why I tried to distinguish between types (a) and (b).

MichaelG.

.

P.S. [James]: You may also like to look at the Huygens 'endless' weight system [which avoids the need for a Maintaining Power device].

So whats wrong with stopping the clock if you must. I thought this was a basic clock design. I never bother stopping anchor escapement weight driven clocks there is usually enough energy in the pendulum to swing past the pallet tips and even if it did not quite clear on the swing the friction collet on the crutch would prevent any harm. Just means you would need to put it back in beat. It's different if you are going for a delicate escapement or you are talking about a large tower clock movement which in the latter case is going to remove the drive for long periods, has a massive pendulum and is awkward to stop and start. With a eight day going weight driven clock with no temperature compensation you are going to adjust the time each time you wind in anyhow.

Regards Martin

This book on the archive might help you with the escapement also some other aspects.

**LINK**

I think it's worth buying the 2 design book I mentioned but when it comes to the weight he suggests setting it to 10 to 15% over that needed to drive the clock. The excess being to account for changes over time.

John

-

Edited By Ajohnw on 20/04/2016 09:53:01

Edited By Ajohnw on 20/04/2016 09:53:30

Surely the pendulum, which is in line with the crutch and therefore the pallet arbor, is going to have to be behind the rear clock plate? So no need to move the barrel to one side.  This actually points up a problem with this design: the rear plate will need to be stood off the back board, and as shown with the long thin plates the pillars are going to be nearly on the vertical axis.  Usually you have a wide back plate so they can be set apart horizontally to clear the pendulum rod.

Huygens endless loop is a good system but really needs a fine chain to implement it which is hard to find.

On pendulums, there is a book by Robert Matthys called Precision Clock Pendulums which is very complete, if you can find or borrow a copy. Or copy the design from an existing clock!

In ME a few years back there was a construction series for the "Repton Clock", written by Stan Bray I believe, and you could buy a material kit including gears and pinions. I wasn't impressed by the series as it included at least one egregious error which would have stumped anyobe following it which I pointed out at the time but was never corrected. I still have the kit which I didn't complete as I got bored with it (though I didn't get caught out by the error!). I was intending to fit a Brocot escapement instead of the anchor and made the new scape wheel needed but didn't complete the pallets.

Edited By John Haine on 20/04/2016 10:12:32

.

For info. ... In Horologium Oscillatorium, Huygens specifies a rope/cord.

MichaelG.

Agreed Martin but if someone inexperienced winds it without stopping it first will they know how to put it back in beat. A simple spring mechanism in the main wheel overcomes the chance of a problem.

Russell.

Edited By Russell Eberhardt on 20/04/2016 11:42:26

I'm not suggesting that maintaining devices are not a good idea but they are not a neccessity on simple clocks. On the typical long case clock they are a rare bird. If you really feel the need then a simple weight lever with a pawl onto the great wheel and a shutter to obscure the winding square would be the most straight forward idea.

regards Martin

I don't see why winding should mean that the pendulum gets in the way as I assume it's mounted on a back plate and the movement mounted on that driving the pendulum in the usual way.

One aspect I have picked up on various places relates to the train. 11 teeth manages to get 2 teeth active with a cycloid / lantern pinion arrangement. As that is not a nice number 12 is generally used. Or more of course if needed. Going on that 10 is a bad choice and doesn't gain anything over say the usual 8.

I think that old book I linked to gives sufficient information to work out the escapement arrangement to get a given pendulum swing. It suggests 7 1/2 teeth on a dead beat and the drawings it shows look about right. It also covers others.

I wont mention what the book that gives 2 designs uses as that would be unfair on the person that wrote the book.

Harrison used a remotoire on his weight driven regulators using springs. It adds to the complications done that way as the power to the pendulum changes over a 15 sec interval - hence his cheeks each side of the suspension spring which were adjustable. A set of feeler gauges might be the best way of obtaining the steel for the spring as changing the thickness has an effect but more limited than the cheeks.

As a fusee was mention earlier I think, Here's how that is done

http://mb.nawcc.org/showthread.php?67822-Building-a-Pinwheel-Skeleton-Clock/page3

I also posted it in another thread. It's pretty obvious that to be any good it needs to be tailored to the spring that is used.

John

-

Edited By Ajohnw on 20/04/2016 12:56:05

Edited By Ajohnw on 20/04/2016 12:56:29

Apologies of I didn't make myself clear. My point was that James' sketch shows all the wheels having their pivots in line on the vertical centreline of the clock, with long thin vertical plates. The pallet arbor is on the same line and so the pendulum neutral axis has to be also. It will be therefore hard to mount stand-off pillars on the top and bottom of the plates that can space the movement off the back board to clear the pendulum, which don't themselves get in the pendulum's way.

To clarify on Harrison's remontoire, the suspension cheeks are not used to compensate for the variation in driving torque from the remontoire, but in conjunction with the grasshopper characteristics in a complicated balance between (near) circular deviation compensation and compensation for barometric effects. The final adjustment of Clock B at the RGO required quite a lot of experimentation with cheek radius and spring thickness/material to get it right.

His remontoire has two functions: one to keep the driving force as constant as possible; the other to allow the scape wheel to recoil without the pendulum having to drive the whole of the train.

If "maintaining power" is desired to keep the clock going while winding, with modern components it should be possible to arrange a small spiral spring in the train, which gets wound by the weight torque and transfers it to the rest of the clock, with roller clutches to lock the "input" of the spring during winding.

I cannot help thinking we are getting a little overcomplicated without perhaps understanding the reasons that the various mechanisms were added to particular clocks. Apologies to all those cleverer than me who know all this already.

Perhaps I could sum up some of the main points whilst in no way professing to be an expert but with some history of clock making and a long time interest in the practical and theoretical aspects of clockmaking.

The period of Pendulums as oscillators can change by reason of thermal expansion of suspension and rod, barometric pressure and variation in the amplitude of the swing (circular error).

For common clocks none of this was compensated for.

The biggest factor in the variation of the period is temperature. For a weight driven clock there is no point in messing around with remontoires unless you have sorted the temperature compensation.

The next biggest issue is circular error. Bigger swings take longer times. That's why Harrison added cheeks to the suspension spring on the RAS regulator. The reason for using a remontoire in the power train of a weight driven clock is it isolates the force delivered to he excapement and thus to the oscillator from the varying frictional forces in the train and tidal variation in gravity and therefor further reduces circular error. (The pendulum will still be affected directly by changes in g).

For Spring driven clocks the driving force reduces as the spring winds down. Fusees are just variable gearing to even out this force reduction.

Remontoires are force regulators and rely either on a wound spring to couple the main driving force to the escapement by consistantly rewinding that spring by a small fraction to keep it at a near constant tension (torsion?). The same can be done by a weight on the end of a lever which is lifted by a small amount at each rewind. In tower clocks remontoirs are important because the forces developed on the large external hands by the weather can create large vaiations in the available driving force for the clock. Remontoires isolate the escapement from the power train.

Rewinding gear is not a remontoire in this sense, it is just there to rewind the clock.

Unless you are attempting to make a very accurate timepiece you can keep things as simple as possible and forget most if not all of these issues. I would go as far as to say that if it's got an anchor escapement and no temperature compensated pendulum you are wasting your time adding anything else unless it's just for the fun of making it which of course should not be denied.

I hope this is interesting and no one feels lectured at cos I don't mean to do that.

best regards Martin

Quite!

Just a minor point: Remontoire is French for a device that re-ascends or re-lifts so perhaps it is more correctly applied to a rewinding device than a force regulator. But then our use of non English terms is often at variance with the original meaning.

Russell

I think that the idea of moving the weight line away from the pendulum is to stop the weight and pendulum swinging together in sympathy as they both reach the same length This can be detrimental to the stability of the clock itself Please don't ask me to explain I just know it can happen

Roy