precision regulator runs slightly faster in extreme heat... ?!?

Discussion in 'General Clock Discussions' started by bruce linde, Sep 11, 2017.

  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  1. bruce linde

    bruce linde Technical Admin
    NAWCC Member

    Nov 13, 2011
    1,833
    6
    38
    Country Flag:
    Region Flag:
    my clock mentor has a lange regulator, c1840... extremely precise. it hangs across from his downstairs shop, mostly sheltered from the outside.

    he went through it 40 days before the big heat wave that struck the sf bay area, doing a full service. for the 40 days prior to the heat wave, it was dead on... it did not gain or lose a second.

    when the heat wave hit, the temp downstairs broke out of its usual 68-72 and hit 75... and the lange gained 3 seconds over 6 days. it seems to stay stable when the temp is 69-73 degrees.

    he thinks the center post is zinc, and the side rods are steel... and all five compensate.

    what would make it speed up? shouldn't it lose time?





    [​IMG][​IMG][​IMG][​IMG]

    [​IMG]
     

    Attached Files:

    • 9.jpg
      File size:
      328.4 KB
      Views:
      1,181
    • 10.jpg
      File size:
      328.5 KB
      Views:
      222
    • 11.jpg
      File size:
      319.1 KB
      Views:
      220
    • 12.jpg
      File size:
      315.1 KB
      Views:
      220
  2. glenhead

    glenhead Registered User
    NAWCC Member

    Nov 15, 2009
    815
    11
    18
    Telecom Engineer
    Williamson County, Texas
    Country Flag:
    Region Flag:
    Guesses, in decreasing order of likelihood:

    1. Gremlins.

    2. Zinc has a coefficient of thermal expansion that is three times that of steel, unless it's Austenitic stainless steel in which case it's only double. The zinc rod is expanding up, and all the steel rods are expanding down. There are twice as many steel rods as zinc, and they're all a bit longer, but there's not enough to equal three times the length. Unless the expansion rates were carefully matched based on the specific alloys used, the pendulum that long is going to get about a hundred-thousandth of an inch shorter per degree Fahrenheit as the temperature goes up. The extra three degrees may have been enough to make it fall outside the "normal" range, though that seems to be a bit odd.

    3. Warmer air is thinner, so there's less air resistance.

    4. The increased temperature resulted in the area rising by several hundred feet, and nobody noticed. That caused a decrease in the effect of gravity, making it run faster.

    As I say, I expect gremlins are the culprits, although zinc-vs-steel comes in a close second.

    Glen
     
  3. Dr. Jon

    Dr. Jon Registered User
    NAWCC Member

    Dec 14, 2001
    5,015
    15
    38
    Aerospace Engineer
    New Hampshire
    Country Flag:
    Region Flag:
    My guesses are:

    1) Compensation is not quite right, slightly overcompensating and/or
    2) Thermal gradients- air is a varying temperatures and pendulum temperature is not uniform. Reifler noted this and designed a special pendulum to account for this
     
  4. burt

    burt Registered User

    Sep 5, 2008
    486
    12
    18
    Male
    McKees Rocks,Pa.
    Country Flag:
    Region Flag:
    Having some experience with my own regulator I noticed you didn't mention anything about the barometric pressure during your running fast period nor do you illustrate any compensating feature the clock has for this error. As answered there can be many factors involved in the "error" you describe but my guess the air pressure is a good suspect.
     
  5. Dr. Jon

    Dr. Jon Registered User
    NAWCC Member

    Dec 14, 2001
    5,015
    15
    38
    Aerospace Engineer
    New Hampshire
    Country Flag:
    Region Flag:
    Air pressure is very tricky. Since it was not mentioned, or measured, it is a great suspect.
    I have done some modeling of complex pendula and found that air resistance is very small factor in pendulum energy loss. Detailed theory from Airy and successors is that resistance loss in itself does not change the rate and my analysis suggests that the change is not enough to affect amplitude enough to cause appreciable circular error. (my analysis to too crude to be very assertive) I believe that the buoyancy is the principal rate changing factor.
     
  6. bruce linde

    bruce linde Technical Admin
    NAWCC Member

    Nov 13, 2011
    1,833
    6
    38
    Country Flag:
    Region Flag:
    the clock is in an internal hallway. it ran for 40 days with losing or gaining a second... with wonkiness mapping (apparently) directly to the spike in temperature and then returning to normal when the temperature did. why would you think barometric pressure rather than perhaps metal fatigue or age (or pendulum design) causing uncalibrated over-compensation?

    i did find this chart on one of the weather services... seems to support temp over pressure? pressure was lower on the hottest day, but didn't vary as much moving forward into when the clock re-stabilized.

    [​IMG]
     
  7. burt

    burt Registered User

    Sep 5, 2008
    486
    12
    18
    Male
    McKees Rocks,Pa.
    Country Flag:
    Region Flag:
    #7 burt, Sep 14, 2017
    Last edited: Sep 14, 2017
    I'm myself am only guessing? Perhaps I should have added a "also" to the end of my post? I'm a researcher with an eye for history and not a precision clock expert. I was thinking out most of my own experiences when I (we) question why a clock does this or that. Perhaps your experience wasn't even related to the temperature? Sometimes even vibration caused my any number of factors is the culprit? Will that account for the entire error the clock experienced (?) Probably not as those errors are not that great as the clock experienced. Another factor is that a rise in temperature will decrease the density of the air and a rise in humidity will do the same and add to the speed of the clock. "The effects of change in barometric pressure may be of a short term nature". (Precision Pendulum Clocks by Derek Roberts) As you reported the clock returned to "normal" when the temperature did.
    How could any deterioration in the metal cause your problem and then correct itself later? Anyway that's why I guessed and only wished to add another set of factors to consider in analyzing a problem that now has seemed to go away. Those are the ones that are the most difficult to sort out.
     
  8. bruce linde

    bruce linde Technical Admin
    NAWCC Member

    Nov 13, 2011
    1,833
    6
    38
    Country Flag:
    Region Flag:
    burt - totally appreciate the time and thought that went into your post... and i'm the first to admit i'm a newbie and that i'm making this up as i go. my comments about metal were in response to dr. jon's and glenhead's comments about over-compensation... i was just thinking out loud what affect 170+ years might have on the compensating rods, and surmising that it might throw off compensatory (!) accuracy.

    i'm surprised the temperature even moved in the interior hallway where the clock resides... i would have thought it was fairly well-insulated against the elements, but apparently not.

    i think your comment that 'a rise in temperature will decrease the density of the air and a rise in humidity will do the same and add to the speed of the clock' is spot on... now i have to see if i can find numbers for mapping humidity during the recent heat spell(s).
     
  9. marylander

    marylander Registered User

    Sep 9, 2008
    824
    3
    18
    Electrical Design Engineer
    Maryland
    Country Flag:
    Region Flag:
    Hello, I am not in the lever to discuss such accuracy clock. This is just a thought.
    You all discussed the thermal expansion of pendulum material. Do we also need to take the air density at different temperature into account? When higher temperature, the pendulum become longer, but the air density will become thinner to reduce the air drag. The effect of the air density is against that of the effect of the pendulum material. Even though, the air density change is ever so small, it still need to be considered since you are dealing with high accuracy clock.
    Ming
     
  10. R. Croswell

    R. Croswell Registered User
    NAWCC Member

    Apr 4, 2006
    6,712
    32
    48
    Male
    Trappe, Md.
    Country Flag:
    Region Flag:
    Clocks like this are above my pay grade but it sounds like the temperature compensation isn't. Would be interesting to see how it reacts over a wider range of temperatures.

    RC
     
  11. Ralph

    Ralph Registered User
    NAWCC Member Sponsor Donor

    Jan 22, 2002
    4,408
    8
    38
    Country Flag:
    Another thought... could the pendulum be suffering from stiction. Are all the elements free, where they should be, to move. It was one of the problems with grid iron pendulums which caused the move to mercury pendulums for some.

    Ralph
     
  12. burt

    burt Registered User

    Sep 5, 2008
    486
    12
    18
    Male
    McKees Rocks,Pa.
    Country Flag:
    Region Flag:
    Something to think about? The regulator gained 3 seconds in 6 days. That works out to be +3 seconds in 518,400 or over a half million beats. That's from a purely mechanical instrument, equipped with no barometric compensation, made how many years ago, that's hanging on a wall in a residential home. If you really want to drive yourself nuts attach a Micro Set timer to it and watch the deviations!
     
  13. shutterbug

    shutterbug Super Moderator
    NAWCC Member

    Oct 19, 2005
    34,632
    43
    48
    Male
    Self employed interpreter/clock repairer
    Iowa
    Country Flag:
    Region Flag:
    I'm guessing, but old regulators used to use actual mercury in their pendulums. It's possible that the original pendulum was replaced after a mishap with the wrong type (non-compensating).
     
  14. bruce linde

    bruce linde Technical Admin
    NAWCC Member

    Nov 13, 2011
    1,833
    6
    38
    Country Flag:
    Region Flag:

    all of the examples of langes we found during research had similar pendulums.

    this clock is c1840... when did mercury pendulums happen?
     
  15. shutterbug

    shutterbug Super Moderator
    NAWCC Member

    Oct 19, 2005
    34,632
    43
    48
    Male
    Self employed interpreter/clock repairer
    Iowa
    Country Flag:
    Region Flag:
    Perhaps before that. Here's one example.
     

    Attached Files:

  16. burt

    burt Registered User

    Sep 5, 2008
    486
    12
    18
    Male
    McKees Rocks,Pa.
    Country Flag:
    Region Flag:
    According to Derek Roberts in his book "Precision Pendulum Clocks" Chapter 5 "George Graham commenced c.1715,to measure the coefficients of expansion of the various metals that might be used in the construction of a pendulum so that they be automatically compensated for any temperature changes." "In 1721 he resumed his trials testing various other materials which may be used as a pendulum bob,including mercury."Illustrated in the book is a Graham's mercury compensated pendulum clock circa. 1727. Robert's goes on to write that testing took place over a three year period and the conclusion was that it worked to some degree. The early problems was with glass containers to contain the mercury and it was determined that metal containers were superior in conducting the actual temperature change. Further experimentation with glass containers of two or more and utilizing thinner glass material greatly increased the surface area of the mercury and speed up its thermal compensation. Frodsham and Dent both followed in the 1830's -1840's using mercury compensated pendulums. By 1850 it was written,"The compensation-pendulum now most generally employed for the best regulators,is the mercurial; and the glass jar has in most cases given way to a steel jar." I could go on but I think that answers the question.
     
  17. burt

    burt Registered User

    Sep 5, 2008
    486
    12
    18
    Male
    McKees Rocks,Pa.
    Country Flag:
    Region Flag:
    It's between the early first uses of mercury to compensate for thermal expansion and its return to popular use by Frodsham and Dent that the Gridiron Pendulum was invented. The Gridiron Pendulum was the second major attempt to compensate for temperature errors in precision timekeeping. John Harrison in about 1726 developed the gridiron pendulum by using rods of steel and brass in a successful combination to "reduce" temperature changes effecting a clocks ability to keep precision time. The gridiron pendulum clock became the most popular for use in England's observatories during the 18th century but by the 19th century the mercurial pendulum again had gained the "ascendancy" until eventually being replaced with the invention on Invar.

    Shutterbug: That is a beautiful regulator!
     
  18. bruce linde

    bruce linde Technical Admin
    NAWCC Member

    Nov 13, 2011
    1,833
    6
    38
    Country Flag:
    Region Flag:
    why does this stuff make me so happy? thx for sharing the info!!!!

    SB - what year is your clock?
     
  19. Ralph

    Ralph Registered User
    NAWCC Member Sponsor Donor

    Jan 22, 2002
    4,408
    8
    38
    Country Flag:
    Sb, it's apples and oranges, I think. I believe Bruce is talking about a precison regulator, which I suspect has an astro dial configuration, jeweling, high pinion counts, etc...

    They are used in scientific setings, labs observatories...

    Ralph
     
  20. bruce linde

    bruce linde Technical Admin
    NAWCC Member

    Nov 13, 2011
    1,833
    6
    38
    Country Flag:
    Region Flag:

    the clock i'm talking about is a gorgeous (what i would call) precision regulator with a non-mercury compensating (or at least supposed to be) pendulum. you can see more photos of it in this thread: https://goo.gl/AJEgEz

    40 days without gaining or losing a second seems pretty impressive to me. my clock mentor appreciates the feedback and will be servicing the pendulum soon to see if it makes a differences; he's been through the movement a few times, but realized he's never worked on the pendulum.
     
  21. Ralph

    Ralph Registered User
    NAWCC Member Sponsor Donor

    Jan 22, 2002
    4,408
    8
    38
    Country Flag:
    That's amazing, if the clock was in fact made for Ohm, as the note in the other thread suggests. George Simon Ohm, German physicist. Wow.

    These precision clocks and chronometers often have incredible provenance.

    Ralph
     
  22. rogerj

    rogerj Registered User

    Dec 21, 2014
    115
    0
    16
    retired broadcast engineer
    Cornwall
    Country Flag:
    Quote " it was dead on... it did not gain or lose a second." for 40 days. A remarkable clock. Is that the same as saying at no time was it ever wrong - even by a second ?
    What happened after the period when it gained 3 seconds...did it become perfect again of loose 3 seconds when the temp returned to normal ?
    From my own obsessive monitoring of a home made clock I know that that over a suitably long period the offset error can return to the same point without giving a clue as to what's happened in between unless a Microset recorder is running (or a Raspberry pi in my case)
    So if the 3 second offset in this case is permanent I would think stiction or some other mechanical settlement the most likely cause. But how that clock performs as well as you say is a mystery given it has no barometric compensation.
    Here's a graph of the September monitoring of my own clock showing the gaining period around the 11th caused by a week of sustained low pressure. It cannot be ignored !
    [​IMG](graphed area is +/- 0.5 secs)


     
  23. shutterbug

    shutterbug Super Moderator
    NAWCC Member

    Oct 19, 2005
    34,632
    43
    48
    Male
    Self employed interpreter/clock repairer
    Iowa
    Country Flag:
    Region Flag:
    I'm not sure, Bruce. Pre-Civil War, I assume. Still running great and keeping great time :)
     
Loading...
Similar Threads - precision regulator runs Forum Date
Any Interest in the study of high precision wall and floor clocks General Clock Discussions Jul 7, 2017
help identify gorgeous precision regulator... lange? is possible? General Clock Discussions Oct 2, 2015
Anyone heard of a Bausch, Lomb, Saegmuller Precision Regulator? General Clock Discussions Jun 16, 2014
Tekippe Precision Regulator General Clock Discussions Oct 28, 2011
Precision Jeweler's Regulator London - Info? General Clock Discussions Feb 9, 2008

Share This Page