# Salomon Coster clock

Discussion in 'Clock Construction' started by Llamazares-Gil, Mar 6, 2015.

1. ### Llamazares-Gil Registered User

Mar 16, 2012
18
0
1
Zaragoza (Spain)
Country Flag:
#1
Last edited by a moderator: Dec 31, 2017
Hi everyone.
With this post I want to present the last clock I've done, it's "El Primero" (the first). Of course it's not a Zenith watch, is the very first pendulum clock ever built.

This clock was conceived by Christiaan Huygens, a prominent Dutch mathematician and scientist born in the Hague (Dutch Republic). Huygens through mathematical calculations determined that, for an accurate clock, his pendulum should follow a cycloidal path in its oscillations and not circular one. The later is what happens when a pendulum is suspended to a fixed point without any other mechanical contrivance. As a consequence of this "circular swing" of the pendulum, circular error comes. In other words the time of the swing varies in proportion to it's arc (no iso-cronyism).

In a clock with a fixed point anchored suspension, when the power applied to train descends, the pendulum`s swing angle descends too. This results that as the energy of the spring decreases over time, the clock varies his rate in the same proportion. This effect is more exaggerated in spring driven clocks rather than in weight driven ones. But in the laters also occurs, since the uneven transmission of energy from the barrel To escape due to imperfections in the gears, also causes loss of isochronyism, and thus loss of accuracy.

To force the suspension to follow the cycloidal path, Huygens developed a system that consisted of a pair of folded sheets called "cycloidal cheeks" of which hung the silk thread pendulum suspension. The curved cheeks shorten the length of the pendulum and counter circular error. In this way, is achieved that the suspension becomes "isochronous". In other words, the period of the pendulum is not dependent on the angle of oscillation.

Without doubt there is a mathematical explanation for all this matters, but that is not scope of this post.

Huygens was a very versatile man. He was mathematician, scientist, astronomer, physicist, horologist and probabilist, but not clockmaker (may not be all perfect). So to carry out his invention he needed a craftsman who built it. He made a prototype by the end of 1656 and in 1657 he contracted the construction of his designs to Salomon Coster a Dutch clockmaker of the Hague, who in this year, was the first clockmaker to make a pendulum clock.

From this moment such watches are known as "Coster clock", and at present all of theese clocks manufactured under Huygen`s license are in museums (the oldest known Huygens-style pendulum clock is dated 1657 and can be seen at the Museum Boerhaave in Leiden) or in private collections.

Back to the present I would like to comment my experience in the construction of this singular clock.
I followed the plans made by C.J. Thorne LBHI and available on the web. I have made changes in what in my judgment improved the aesthetic of the clock. Of course this is questionable.
I'm not versed in the construction of such types of clocks, in fact is the first time I do one whith crown and verge. The first thing that surprised me is the amount of energy they need these types of escapes. Barrel size is fairly large, and the mainspring that corresponds to it is the largest available in the standard supplies. As be imagined, the force that must be performed in order to winding the clock is considerable.
In this clock everything is strong and oversized. Some time ago I restore a half-second pendulum clock of the type "Vienna Regulator" with shafts pivots of the leg size of a mosquito, this clock is located on the opposite side in terms of mechanical concept. You could say that this is a "muscle clock".

The verge (pallets and arbor) is by far, the hardest part of making from all the clock and I suspect that is not properly designed, according to plans. I'll explain. When the clock is in running order, the "tic" corresponding to the palette with short arbor to pivot, rings loud and clear, but he "tac" of the palet in the side of the cruth bounces. It sounds like "trrrac", soft but appreciably. I'm not sure if there is a problem of lack of diameter in the arbor or that, it is too hardened, or simply that in such clocks this is normal.

The back plate engravings are handmade and are a personal contribution as well as the cycloidal cheeks potences, which simulate the wings of a dragonfly. The brand signature is made with pantograph. Make letters engraved by hand is extremely difficult, at least for me.

Regarding its performance is early to draw conclusions, but at first glance, it seems that is far from being a regulator, although it was a major advance over previous clocks with foliot as a system of measuring time.

I have pending perform a provisional anchorage for the pendulum but without the cycloidal cheeks, in order to compare the behavior of the clock with or without the isochronous system. When I carried all measurements I will expose the conclusions in another post.

The next task is to make the case, the hands and the dial. But of course............... that will be another story.

Regards.

Alfonso Llamazares.

2. ### gmorse Registered User NAWCC Member

Jan 7, 2011
10,640
1,272
113
Male
Retired from Xerox
Breamore, Hampshire, UK
Country Flag:
Region Flag:
#2
Hi Alfonso,

That is most impressive!

Tell me, are the verge flags made separately from the arbor or in one piece with it, what angle did you set between them, and how did you make the cheeks to their cycloidal profile?

Regards,

Graham

3. ### Llamazares-Gil Registered User

Mar 16, 2012
18
0
1
Zaragoza (Spain)
Country Flag:
#3
Hi Graham.

The verge flags and the arbor are one piece. The entire assembly is machined flat, then the central shaft portion is heated and then rotated until achieving 37 degrees between the flags. As I commented in the post It is the most difficult part of making in the entire clock.

The cheeks are made as follows: first, with the measurements of the planes make a template in brass as accurate as possible. then make in hardwood a concave mold with the same dimensions, and against this, hammering the flat strips that will form the cheeks with nylon hammer. At all times is necessary to stop the hammering and test the profile with the template. Finally filing up to eliminate any mark of the hammer.

I hope the explanation is helpful to you.

4. ### gmorse Registered User NAWCC Member

Jan 7, 2011
10,640
1,272
113
Male
Retired from Xerox
Breamore, Hampshire, UK
Country Flag:
Region Flag:
#4
Hi Alfonso,

Thanks, that is indeed helpful. I was interested to note that the flag angle is only 37 degrees, much narrower than later clocks, which seem to work best at around 70 to 75 degrees.

What type of case are you planning?

Regards,

Graham

5. ### Tinker Dwight Registered User

Oct 11, 2010
13,666
63
0
Calif. USA
#5
The clock is fantastic. You should be proud, I know I would be
with such a clock.
To balance the drops, you'd move the crown side ways.
To close the drops, you'd raise the crown. If you don't
have enough range without locking, you'd change the
angle ( twist ).
It sounds like you need a little of each.
Of course, the silk threads are some times used
for humidity scales.
Tinker Dwight

6. ### tok-tokkie Registered User

Nov 25, 2010
310
1
18
Cape Town, South Africa
Country Flag:
#6
Lovely clock you have made. I really like the dragonflies you have added to it.
I had not realised that the initial Huygens clocks were spring driven. I wrongly believed that the two with long pendulums and weight drive were his original pendulum clocks.

7. ### Llamazares-Gil Registered User

Mar 16, 2012
18
0
1
Zaragoza (Spain)
Country Flag:
#7
Hi Graham.
I do not know exactly the reason why CJ Thorne has decided to design this clock with this angle flags, but I can get an idea. Looking at the book "practical clock escapements" from Laurie Penman, in chapter 2 "the verge escapement" we can read.
- One of the important factor in deciding the shape of a verge arbor and pallets is: the flags may be at any angel that the maker has found convenient -.

When deciding the angle of the flags must take into account the advantages and disadvantages. Wide flag angles give lower pendulum swing angles therefore low circular error as well as it makes the escapement easier to manufacture and less prone to "skiping" as the flags wear. On the other hand there is a high upward presure to the flags when the recoil begins (not suitable for for clocks with knife-edge suspension), and prone to premature wearing downs. With low flalgs angles the opposite occurs.

However there are intermediate options to make a clock with low angles of flags and relatively low angles of pendulum, with the use of a crutch whose axis of oscillation (fulcrum) is below the axis of the pendulum.
This is the case of this clock, which with 37 degrees of flags, the pendulum swings at 31 degrees. This pendulum angle is equivalent to an angle of 85 degrees of flags in a clock whose pendulum is directly attached to the arbor of the verge (without crutch).
I do not know, if you get to understand my explanation. My English is not all good.

I want to make the case as much like the original as possible. On the plans there are also instructions and measures to realize it.

regards.

8. ### gmorse Registered User NAWCC Member

Jan 7, 2011
10,640
1,272
113
Male
Retired from Xerox
Breamore, Hampshire, UK
Country Flag:
Region Flag:
#8
Hi Alfonso,

Thanks, yes, your explanation is quite clear.

My figure of 70 to 75 degrees was indeed in the context of knife-edge suspensions, and thus not applicable to your clock.

Regards,

Graham

9. ### Tinker Dwight Registered User

Oct 11, 2010
13,666
63
0
Calif. USA
#9
If the bob weight follows a cycloidial path, it will take the same
amount of time ( isochronal ).
As the pendulum swings, instead of the rate slowing
from circular error, the spring shortens, making the bob
want to move faster.
If shaped right, the circular error is canceled.
Tinker Dwight

10. ### cmnewcomer Registered User

Mar 24, 2009
48
1
8
#10
Very impressive Alfonso. Thanks for sharing.

Best Regards.

Carl

11. ### Llamazares-Gil Registered User

Mar 16, 2012
18
0
1
Zaragoza (Spain)
Country Flag:
#11
Hi Tinker.
I checked again all parameters concerning the escapement and are all correct. However although I vary them, the bounce in the long axis flag remains.
It seems to me that is not a problem of geometry in the escapement.

regards.

12. ### ccwk Registered User NAWCC Member

Jan 27, 2011
17
0
1
Country Flag:
Region Flag:
#12
Looks like a great project and I look forward to seeing the case and finished product.
Thanks for sharing, keep up the great work
Cheers Conwae

13. ### Ralph Registered User NAWCC MemberSponsor

Jan 22, 2002
4,700
75
48
Country Flag:
#13
Very nice. Thank you for posting it to this forum. It's great to see the skill some of the members possess.

Will we be seeing #1 through #3? ....and looking forward to#5

Cheers, Ralph

Feb 19, 2005
3,899
48
48
Cabinetmaker