• The NAWCC Museum and Library & Research Center are currently open. Please check the Visiting Schedule for Days and Hours at the bottom of the Visit Page.

Atmos Temperature Compensation

Dr. Jon

Moderator
NAWCC Member
Dec 14, 2001
6,195
671
113
New Hampshire
Country
Region
I have found that the suspension on the Atmos is elinvar. As I understand elinvar, its spring constant is very small. The balance seems to be brass and I think its expansion leads to temperature error. What am I missing? Does the balance have temperature compensation?
 

Burkhard Rasch

Registered User
NAWCC Member
Jun 1, 2007
4,885
181
63
64
Twistringen
Country
Region
I´m sure John Hubby will give a more comprehensive answer to Your interesting question,but AFAIK the main temperature error in torsion clocks is caused by the change of the modulus of elasticity of the suspension spring with changing temperatures.This error overtops the error caused by the expansion/contraction of the pendulum ring by factor 10, so-again AFAIK-there´s no temperature compensation built in the pendulum rim of my Atmos.There are indeed temperature compensating pendulums in anniversaries,but their influence is outruled when using a Horolovar compensating suspension spring. HTH with my 2cents worth,could be wrong though.
Burkhard
 
Last edited:

John Hubby

Senior Administrator Emeritus
Staff member
NAWCC Star Fellow
NAWCC Life Member
Sep 7, 2000
12,289
231
63
The Woodlands, TX
Country
Region
Dr. Jon;532318 said:
I have found that the suspension on the Atmos is elinvar. As I understand elinvar, its spring constant is very small. The balance seems to be brass and I think its expansion leads to temperature error. What am I missing? Does the balance have temperature compensation?
Dr. Jon, I can explain what happens with the Atmos. The balances do "not" have temperature compensation, what is compensated is the Elinvar suspension spring. Here's the story:

First of all, what is unique about Elivar, Nivarox, NiSpan C, and similar alloys is that their modulus of elasticity (Young's modulus or spring constant) is very stable over a wide temperature range, not that is it small. All three materials are used for torsion clock suspension springs, watch hairsprings, and other applications requiring constant modulus for accuracy. All of these also have nearly zero coefficient of expansion over the same range of temperatures in which the modulus is relatively constant, which in turn is especially important for precision hairspring applications. The inventor of Elinvar, Charles-Edouard Guillaume, recognized this in the development of the Guillaume overcoil hairspring balance for precision watches in the early 1920's.

Among the three alloys, Elinvar is unique in that its modulus can be precisely changed by special treatment to obtain either increasing or decreasing modulus behavior, whereas the other two do not exhibit this to nearly the same degree. Also, mechanical processing of Elinvar is somewhat more difficult than the other two; that is why Nivarox was used in Europe from the late 1930's and NiSpan C was used by Hamilton for their chronometers in WWII. They gained the needed performance for hairspring applications without needing special treatment.

However: J. L. Reutter and later LeCoultre determined that because of the relatively large mass of the pendulum ring or disc used for Atmos clocks coupled with the very slow half minute beat, that Elinvar provided what was needed to compensate for thermal expansion of the pendulum. You are right on that count, the pendulum does expand measurably between 0 and 35 deg. C and thus changes the center of mass which would nominally slow down the clock as temperature increases or speed it up as temperature decreases. Reutter then developed a treatment of the Elinvar spring material that has been refined by LeCoultre (I have seen this but am not at liberty to reveal what is done), that gives it a slightly increasing modulus with temperature that theoretically exactly matches the thermal expansion and contraction of the pendulum. For that reason an Atmos clock properly regulated is the most accurate mechanical clock made that doesn't require any kind of special escapement or other device to achieve that accuracy.

I hope this will explain how the Atmos gets its accuracy. 400-Day clocks are not nearly the precision instruments that Atmos clocks truly are, thus "IF" there is any effect of pendulum expansion/contraction it is already overshadowed by a multiplicity of other factors including variable mainspring power, escapement irregularities, etc. Thus, Nivarox or NiSpan C are perfectly suitable for suspension springs for these clocks. And, properly regulated a well-set-up 400-Day will keep excellent time (but not at all as good as an Atmos).
 

Burkhard Rasch

Registered User
NAWCC Member
Jun 1, 2007
4,885
181
63
64
Twistringen
Country
Region
Thank You John,I´ve again learned a lot!
Burkhard
 

Dr. Jon

Moderator
NAWCC Member
Dec 14, 2001
6,195
671
113
New Hampshire
Country
Region
Thanks from me too.

I had been aware that post melt processing was important, which is why the makers have no issue with providing the chemistry.

The neat thing to me is that they can get the coefficients to approximate the temperature curve, which is not a straight line, needed to correct for pendulum expansion.
 
Know Your NAWCC Forums Rules!
RULES & GUIDELINES

Find member

Staff online

Forum statistics

Threads
162,326
Messages
1,409,797
Members
84,020
Latest member
jsteifle
Encyclopedia Pages
1,101
Total wiki contributions
2,854
Last edit
Waltham Watches by Clint Geller