NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: mechanical chronometers
From: Geoffrey Kolbe
Date: 2006 May 15, 23:59 -0500
To determine GMT at any given time you need to know two things about a
given chronometer. The first is how fast or slow to GMT the chronometer was
on a certain date, and the second is its daily "rate" - how much time it
gains or looses per day. With these two pieces of information, you should
be able to calculate the actual GMT to some degree of accuracy from the
time given by the chronometer. So, what matters is that the rate at which
the chronometer gains or looses time is constant with time and does not
change - particularly with temperature.
For a chronometer to be certified as such, it has to pass certain tests
which determines the constancy of its daily rate. In the main, the tests
are not designed to see how accurate the chronometer is as a time keeper -
its accuracy as a timekeeper is actually of secondary importance.
In the 1940's for example, the National Physical Laboratory in Teddington,
near London, specified that a chronometer had to pass the following
conditions to be issued with a Class "A" KEW certificate.
The chronometer was tested over a period of 55 days. This was divided into
5 periods of 11 days at five different temperatures. The first period was
at 70°F, the second at 45°F, the third at 70°F again, the forth at 95°F and
the fifth back to 70°F again.
The mean daily rate was determined for each period and it should not exceed
7.5 seconds for any period of the test. This is actually pretty generous as
a condition and reflects the fact that accuracy in keeping time was not
that important. My Ulysses Nardin chronometer, for example, has a mean
daily rate of about +0.66 seconds per day, so a daily rate better than a
tenth of that required is fairly easy to achieve.
A tougher test was that as the temperature was changed from period to
period, the mean daily rate should not change by more than 0.1 seconds per
day per degree Fahrenheit. So, as the temperature was raised or lowered by
25°F, the rate should not change by more than 2.5 seconds per day.
Also, the average amount by which the daily rate changes from day to day
(the mean variation of the daily rate) should not exceed 0.5 seconds during
any of the test periods. That is pretty tough for a mechanical clock,
particularly over the temperature ranges to which the chronometers were
subjected. Most respectable ordinary clocks have a mean variation of around
5 seconds a day. A really good mechanical chronometer can manage better
than 0.2 seconds per day as a mean variation and around 0.02 seconds per
day per °F. A large contribution to the variation in the daily rate of a
chronometer is actually changes in atmospheric pressure, or air density to
be more specific.
To achieve this sort of performance, many design features are present in a
chronometer that are simply too expensive to put into a normal clock.
Special oils are required on different parts of the chronometer to keep it
working sweetly over long periods. This is why the average high street
clock repairer is not qualified to work on chronometers. He (or she) will
not be familiar with the mechanisms in a chronometer through their
experience working on the more usual mechanical clock.
A good book on the subject of chronometers is "The Ship's Chronometer" by
Marvin E. Whitney, published in 1985 by American Watchmakers Institute
Press ISBN 0-918845-08-4. Not cheap though!
Geoffrey Kolbe
At 18:26 15/05/2006, you wrote:
>Geoffrey-
> "It must be understood that a chronometer is not a "clock"."
>
>I'm afraid I only know the modern use of the term, i.e. that a wristwatch
>having
>passed certain ratings is now "certified" as a chronometer, as Rolex does. Can
>you tell me what makes a chronometer not just a clock, in the larger
>sense? And
>what accuracy you expect from a true chronometer?
>
>
>
Border Barrels Ltd, Newcastleton, TD9 0SN, UK
Tel: +44 (0)13873 76253 Fax: +44 (0)13873 76214
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to NavList@fer3.com
To , send email to NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---
From: Geoffrey Kolbe
Date: 2006 May 15, 23:59 -0500
To determine GMT at any given time you need to know two things about a
given chronometer. The first is how fast or slow to GMT the chronometer was
on a certain date, and the second is its daily "rate" - how much time it
gains or looses per day. With these two pieces of information, you should
be able to calculate the actual GMT to some degree of accuracy from the
time given by the chronometer. So, what matters is that the rate at which
the chronometer gains or looses time is constant with time and does not
change - particularly with temperature.
For a chronometer to be certified as such, it has to pass certain tests
which determines the constancy of its daily rate. In the main, the tests
are not designed to see how accurate the chronometer is as a time keeper -
its accuracy as a timekeeper is actually of secondary importance.
In the 1940's for example, the National Physical Laboratory in Teddington,
near London, specified that a chronometer had to pass the following
conditions to be issued with a Class "A" KEW certificate.
The chronometer was tested over a period of 55 days. This was divided into
5 periods of 11 days at five different temperatures. The first period was
at 70°F, the second at 45°F, the third at 70°F again, the forth at 95°F and
the fifth back to 70°F again.
The mean daily rate was determined for each period and it should not exceed
7.5 seconds for any period of the test. This is actually pretty generous as
a condition and reflects the fact that accuracy in keeping time was not
that important. My Ulysses Nardin chronometer, for example, has a mean
daily rate of about +0.66 seconds per day, so a daily rate better than a
tenth of that required is fairly easy to achieve.
A tougher test was that as the temperature was changed from period to
period, the mean daily rate should not change by more than 0.1 seconds per
day per degree Fahrenheit. So, as the temperature was raised or lowered by
25°F, the rate should not change by more than 2.5 seconds per day.
Also, the average amount by which the daily rate changes from day to day
(the mean variation of the daily rate) should not exceed 0.5 seconds during
any of the test periods. That is pretty tough for a mechanical clock,
particularly over the temperature ranges to which the chronometers were
subjected. Most respectable ordinary clocks have a mean variation of around
5 seconds a day. A really good mechanical chronometer can manage better
than 0.2 seconds per day as a mean variation and around 0.02 seconds per
day per °F. A large contribution to the variation in the daily rate of a
chronometer is actually changes in atmospheric pressure, or air density to
be more specific.
To achieve this sort of performance, many design features are present in a
chronometer that are simply too expensive to put into a normal clock.
Special oils are required on different parts of the chronometer to keep it
working sweetly over long periods. This is why the average high street
clock repairer is not qualified to work on chronometers. He (or she) will
not be familiar with the mechanisms in a chronometer through their
experience working on the more usual mechanical clock.
A good book on the subject of chronometers is "The Ship's Chronometer" by
Marvin E. Whitney, published in 1985 by American Watchmakers Institute
Press ISBN 0-918845-08-4. Not cheap though!
Geoffrey Kolbe
At 18:26 15/05/2006, you wrote:
>Geoffrey-
> "It must be understood that a chronometer is not a "clock"."
>
>I'm afraid I only know the modern use of the term, i.e. that a wristwatch
>having
>passed certain ratings is now "certified" as a chronometer, as Rolex does. Can
>you tell me what makes a chronometer not just a clock, in the larger
>sense? And
>what accuracy you expect from a true chronometer?
>
>
>
Border Barrels Ltd, Newcastleton, TD9 0SN, UK
Tel: +44 (0)13873 76253 Fax: +44 (0)13873 76214
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to NavList@fer3.com
To , send email to NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---
