1 second Regulator – state of play

Discussion in 'Clock Construction' started by Phil Burman, Nov 30, 2016.

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

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    #1 Phil Burman, Nov 30, 2016
    Last edited: Nov 30, 2016
    Here are some details of the various components of the 1 second regulator I am working on.

    The suspension is a Strasser style twin spring arrangement with a 10mm invar pendulum rod. The pendulum bob geometry is to be decided but will include Riefler style expansion compensation, using an aluminium/brass tube under the bob and sitting on the regulation nut.

    The wheel train uses involute gears (oh dear) and directly drives the separately mounted hour, minute and second hands. The going train is:

    Wheel 1pinion 1Wheel 2pinion 2Wheel 3pinion 3Wheel 4pinion 4Esc. Whlbeat
    # of teeth9616100501281612016301.0000
    ratio6287.5
    Module110.80.80.80.80.80.8
    PCD96168040102.412.89612.8

    There is no motion works, the hands are mounted directly on the going train.

    The movement is sized (or possibly over sized) for a 16” silvered dial, a rough sketch of the current plan for the finished clock is included.

    A number of significant mistakes in the design have been made but as the primary goal is for a precision clock these errors have not yet been corrected where there is no implication for the precision. I can easily visualise that the clock will never be completed (by me) and that I will spend the rest of my life improving the precision. It’s all about the journey not the destination.

    Any comments, good or bad, are welcome.

    Phil:)


    back plate and suspension.jpg backplate and suspension 1.jpg movement 1.jpg movement.jpg movement2.jpg pendulum arrangement .jpg Strasser suspension.jpg suspension spring.jpg suspension suport.jpg wheel train.JPG case layout sketch.JPG
     
  2. Jim DuBois

    Jim DuBois Registered User
    Donor NAWCC Member

    Jun 14, 2008
    1,223
    9
    38
    Retired from computer program/product/project mana
    Magnolia, TX
    Country Flag:
    Region Flag:
    Looks like some first rate craftsmanship! Nice work! Keep the details coming.
     
  3. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    Is that why regulators look like that? I had always assumed it was so that they were easier to read accurately. Is there an advantage in accuracy by excluding the motion work needed to get all the hands coaxial?
     
  4. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    I think regulators take many forms but the basic requirement is that they provide you with sufficient accuracy to allow you to properly regular other clocks and timing devices. This is most easily achieved with a mechanical clock when it uses a pendulum, is weight driven and has a seconds hand. In addition accuracy is improved when friction is reduced, so one theory for separate hands might be the avoidance of additional friction associated with the complication of driving all hands coaxial.

    Phil:)
     
  5. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    Ah,I should have been more specific, as a collector of 18th century English longcase regulators are all very much like yours..
     
  6. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    #6 Phil Burman, Nov 30, 2016
    Last edited: Nov 30, 2016
    Based on a quick look at Derek Roberts book on "English Precision Pendulum Clocks" some 18th century regulators appear to have co-axial minute and hour hands.

    Phil:)
     
  7. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    This is the sort of thing I think of, which is more in line with yours.

    https://www.bonhams.com/auctions/21931/lot/65/

    Regulators still command a higher price here, though obviously this one is a bit special.
     
  8. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Yes a bit special but priced accordingly.

    Phil:)
     
  9. Allan Wolff

    Allan Wolff Super Moderator
    NAWCC Member

    Mar 17, 2005
    493
    3
    18
    Tulsa, OK
    Country Flag:
    Phil,
    Excellent job!
    One item that I wish I had incorporated in my Strasser clock was a way to adjust the beat other than bending the "crutch" shown in the 3rd photo. Very slight adjustments make a big difference in beat. I am certainly interested in any ideas you have in this area.
    Allan
     
  10. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Hello Allan. for the moment I have included a grub screw, with brass pad, see photo. I also used grub screws in the Skeleton clock to allow adjustment of the pallets relative to each other. It would be nice to have some form of micrometer adjustment. I did think of including a fixed disc with tick marks and a single tick mark on the pallet arm, which I think might work well enough to provide a relative reference point when making adjustments, something to thing about down the road.

    regulator pallet grub screw.jpg skeleton clock pallet grub screws.jpg

    Phil:)
     
  11. jhe.1973

    jhe.1973 Registered User
    Donor NAWCC Member

    Feb 12, 2011
    448
    2
    18
    Arizona, USA
    Country Flag:
    Region Flag:
    Yep, know exactly how you feel! If you are anything like me, building and improving something is far and away more rewarding and educational than reading all the available literature.

    Great workmanship and planning IMHO.

    Thanks for posting and I am anxious to follow your progress to learn what discoveries you make along the way.
     
  12. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    263
    0
    16
    Cape Town, South Africa
    Country Flag:
    I am much impressed by this clock.
    Are the frame plates stainless steel? How thick are they?
    I look forward to reading about making it & solving problems.
    I have spent about 20 years making & refining a regulator - also without motion work but mine has conventional face with center sweep seconds hand. It has given me much pleasure to solve the glitches and mysteries. However mine is not of regulator accuracy - sadly.
    Have started drafting a write up.
     
  13. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    The end plates are aluminium, 0.25” x 11” x 6.8”. I originally bought brass plates to the same dimensions, expensive. So being aware of my scope for mistakes and errors and the fact that this clock is a “design as you go” the probability of the end plates coming out the other end unscathed was/is pretty low, so I decided to start with aluminium end plates (and pillars) at a fraction of the cost. This seems to have been a good decision as I immediately proceeded to make several errors on the second plate. I fixed or ignored the mistakes and carried on. If/when I complete the development version I can go back and build a second clock adapted to incorporate all the things I will have learned during the construction of the original. Alternatively I can’t see any technical reason not to stick with aluminium plates, other than tradition. The plates have brass inserts with bearing races supporting all the wheel arbors.

    Phil:)
     
  14. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Here’s where I’m currently at. I’ve been experimenting with difference bob and drive weight configurations in mild steel (without temperature or barometric compensation) and have arrived at the configuration shown in the photo. The bob weight is 13 lb and the drive weight is 6 lb, the amplitude with this configuration is approximately 1.3 degrees. I’m currently redoing this arrangement in stainless steel and brass with an invar rod and temperature compensation.

    The graph is from my Microset timer and it is a great surprise to me that, without any special preparation - straight out of the box so to speak, it shows that the gain/loss over 48 hours never exceeded an incredible + or - 0.1 seconds. I’ve treble check it and it appears to be correct.

    Phil:)
     

    Attached Files:

  15. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    that's a brilliant result.
     
  16. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    263
    0
    16
    Cape Town, South Africa
    Country Flag:
    The information that the Microset gives is fabulous. I am not referring to the brilliant consistency of your clock but to be able to see how the rate of the pendulum changes on each & every swing.

    On your escape when is the impulse applied to the pendulum? As it passes through zero (center of swing and dead point when pendulum is stationary)?
     
  17. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    I'm quite new to clock construction but yes I agree, if a person is serious about working with the construction of a precision regulator then the Microset with software, environment sensors and GPS module seems to me to be virtually essential, even if you have to beg, borrow or steal to get it.

    Yes the impulse is applied at zero amplitude (centre of swing). I double checked it on the simulation here :

    http://www.clockwatch.de/index.html?html/tec/hem/hal.htm

    From the menu on the left select: Theory – Escapements – Detached - Strasser

    Phil:)
     
  18. jhe.1973

    jhe.1973 Registered User
    Donor NAWCC Member

    Feb 12, 2011
    448
    2
    18
    Arizona, USA
    Country Flag:
    Region Flag:
    I am not offering this as an absolute truth. Simply passing along what I was told years ago as to why regulators had separated hands on their dials.

    When regulators were used for extremely accurate time measuring they were used as a support for various research/experiments such as astronomy. Seconds were usually the most important function of a clock. The experimenters were not necessarily 'clock people' and quickly glancing at the clock to record a specific second was made easier (less chance for confusion) by having separated hands for the functions.

    Some precision clocks with separated hands had motion work for the hour hand and some did not. The motion work made setting the clock simpler by allowing the hour hand to keep in step with the minute hand. Those w/o motion work require the resetting of the hour hand if much change is to be made to the minute hand.

    This brings me to a question for Phil.

    Do you have a plan for how you are going to mount the hour hand? It looks to me that if you mount it on the winding shaft for the barrel, it will rotate along with the winding crank.

    This is not a criticism, I'm just curious.

    I am also impressed by you results w/the Microset!

    :thumb:
     
  19. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Well spotted, Yes I didn't fully comprehend the implications when I when ahead with building the prototype. There is a point that comes when I'm just itching to get start on making parts and stop procrastinating over the design.

    I have seen a regulator somewhere that drove the hour hand on the same axis as the drum arbour. As the drum also includes the maintaining wheel assembly I think the only way to do it is to turn the whole wheel/drum assemble around so that the great wheel with maintaining assembly is at of the movement of the movement. Then take a cannon off the great wheel supported on a bearing in the front plate, with the drum winding arbour through the centre of the cannon. This will require some rejigging of the assembly due to the change in winding direction. I did some rough sketches and it looked like it would work OK. I would be interested comments on this and any other alternatives. However, this modification may be rendered unnecessary as I also have another side issue in that the movement is currently only an 8 day. The final intention was/is a 31 day and to achieve this it looks like I will need to move the drum/maintaining wheel assembly to the side and include and additional wheel and pinion on a new hour hand arbour. This is in order to achieve the 31 day within the available drop.

    I’ve purposely kept the movement of the prototype as simple as possible in order to keep maximum focus on the precision.

    As I'm only a beginner and this is my first one second weight driven movement I don't know if the precision shown in the Microset graph is what I should expect straight off?
     
  20. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    263
    0
    16
    Cape Town, South Africa
    Country Flag:
    #20 tok-tokkie, Feb 23, 2017
    Last edited: Feb 23, 2017
    About the separate dials for hrs/mins/secs. I have an interest in old buildings. I was on a tour of the local Royal Observatory recently. They have a museum which includes a working Reifler regulator from 1910. Ian Glass, a retired astronomer, was our guide. He was instrumental in establishing the museum and has written a book about the observatory (and several others). Before his talk & tour I said to him I was particularly interested in any mechanical clocks. He simply said there were a couple. So I was absolutely delighted to come across the Reifler. It is driven by a 30 second remontoire which is rewound electrically. The remontoire was geared directly to the escape wheel arbor. The train down to the minute and hour hands transmitted only motion - no power was running through them.


    Ian Glass said that to make a really accurate transit sighting (transit is time that a star – or the sun – passes absolutely directly overhead when viewed through an astronomical telescope mounted so it can swing on the line of longitude but it can not swing the other way at all. To measure the time of transit the astronomical regulator is mounted close by so the astronomer can hear the seconds tick. Ian went on to say that they could actually time it to a tenth of a second. But each astronomer had to have his personal tenth second estimation calibrated – so that it was accurate.


    Here is the Wiki page on the Reifler escape. https://en.wikipedia.org/wiki/Riefler_escapement It includes a list of the Reiflers on public display and the local one is not included. I intend asking Ian if i may add it.
    This thread is about the Strasser and Haldimann escpaes. This source http://www.historische-zeitmesser.de/veroeffentlichung/rezensionen/rezension_01_2014.html includes this



    The link to the animation of the Haldimann escape points out that although the Haldiman looks similar to the Reifler & Strasser escapes it is functionally different and superior.
     

    Attached Files:

  21. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Just a short post to show the new bob/pendulum in brass, stainless steel and an invar 36 rod. This has replaced the previous prototype mild steel one and now includes temperature compensation via a stainless steel tube sitting on the regulation nut and supporting the bob at its’ mid-point. Also now included is a 5.5lb stainless steel drive weight.

    Phil:)

    new bob 1.jpg
     
  22. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,680
    28
    0
    Calif. USA
    Cool
    Is it a calculated compensation or have you actually run it at a couple
    different temperatures to see how it is doing?
    Tinker Dwight
     
  23. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    #23 Phil Burman, Mar 7, 2017
    Last edited: Mar 7, 2017
    Both. I originally calculated the length and added 10mm to give me scope to home in based on observation. However due to my inexperience I didn't allow enough time for stabilisation consequently the compensator length appeared to be in error. Furthermore being a novice I also managed to confuse myself due to a coincidence that the room thermostat just happened to have a 1 hour cycle and that the clock had a inherent 1 hour rate variation due presumably to an anomaly with the 1 hour wheel. The upshot of this was that the thermostat variation matched the rate variation so I assume that the rate variation was due to this temperature variation. This gave me a calculated pendulum length of 5 meters in order to accommodate a calculated compensator length of 4 meters.:confused:

    So all of the above took me about a week but yesterday I finished a 24hr run with a maximum length compensator and tonight I am running a 24 hour test with the shortest compensator that fits with the length of thread available on the invar rod.

    Here's the plot from yesterday:

    blue is the pendulum amplitude -black is the rate - green is the room temperature

    PS: ingnore the values in the boxes at the top, I was using the gps module so they are not correct/relevant.
     

    Attached Files:

  24. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,680
    28
    0
    Calif. USA
    A good start. I think I'd do some longer constant temperature runs first
    with about 12 hour soak.
    Then you can start looking at rate of change stuff.
    Tinker Dwight
     
  25. tok-tokkie

    tok-tokkie Registered User

    Nov 25, 2010
    263
    0
    16
    Cape Town, South Africa
    Country Flag:
    What do you conclude from those traces?
    The compensator is the maximum possible so the compensation is expected to be excessive?
    As the temperature dropped the period of the pendulum also dropped - which corresponds to over compensation?
    But then what happened? The period reduced in a spike then stabilised.
    In that last stable section the rate varied from 0.000 020 to 0.000 026 sec. So the range was 0.000 006sec. That is 0.5sec/day.
    I have a clock where the daily variation is worse than that. But I have no idea what is usual. What daily change in rate with the temperature being 'reasonably' stable (I don't mean US style temperature control) would you anticipate.
    I recognize that this is a path you are still discovering.
     
  26. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    I think you are right, the trouble is I get too impatient, although I seem to be so far off at the moment that I would like to see a result with under compensation, then at least I know where I am and that the correct compensation is somewhere in between the two.

    The current run with the shortest compensator (not yet finished) appears to also be over compensated, so my plan is to make a small clamp for the invar rod and put it directly under the centre support of the bob, i.e. zero compensation, then I will at least know if the ball is actually in the park.:clap:
     
  27. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    #27 Phil Burman, Mar 8, 2017
    Last edited: Mar 8, 2017
    Here's the plot for the shorter compensator. There seems to be a lot going on but one key point maybe that for the last part of the curve the rate recovers back to the starting point, which I think indicates less over compensation than the previous plot (or possibly zero). The somewhat massive initial drop in rate is surprising but I think it may be due to the pendulum being separately mounted on an alumimium backboard which is in intimate contact with the concrete wall, whereas the movement endplates are in free air, this will result in a differential expansion as the temperature changes. Also the backboard is bolted to the wall both top and bottom. If the bottom bolt predominates then as the temperature falls the pendulum will move down and the crutch will move up. This will change the degree of deflection of the upper suspension spring resulting in a change in input energy to the pendulum.

    Anyway I have now isolated the backboard from the wall with spaces and ensured that the backboard is firmly fixed to the wall by the top bolts, so that direction of movement of the board with temperature is matched by the end plates. We will see, it's running now.

    Phil:)

    PS: you have to remember that the recorded temperature is the instantaneous air temperature and may bear little relationship to the temperature of the various components of the clock.
     

    Attached Files:

  28. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Here's the 16" dial with walnut frame. The dial needs some more work on the graining before silvering. There's a lot of separate steps in the engraving operation, one mistake and it's start all over again, as the reverse side of the dial can testify. I may redo it at some later stage but large lumps of brass can get expensive.

    16 inch dial 3.jpg

    Phil:)
     
  29. Allan Wolff

    Allan Wolff Super Moderator
    NAWCC Member

    Mar 17, 2005
    493
    3
    18
    Tulsa, OK
    Country Flag:
    Phil,
    Excellent job on the dial; very intricate. I can't wait to see it when it is silvered.
    Allan
     
  30. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Thanks Allan, at the moment I'm having trouble getting the right surface finish. The various engravings seem to affect how the wet and dry performs and it all looks a bit streaky. I did have to use a more course paper because of a number of scratches, maybe I just need to persevere with the 600 paper.

    Phil:)
     
  31. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    How are you graining it, vertically?

    I use 800 for graining, though usually I'm restoring and do a lot with 1200 to clean and leave the original graining with with some tarnish so that it doesn't look too bling on an old clock.

    Will be interesting to see your finished dial, I've never worked on a new one.
     
  32. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    I'm graining it circumferentially, pivoted about the centre hole. It's now clear that I should have taken more care regarding scratches on what was originally supposed to be the back side, and possibly I should have grained it before engraving, or maybe not. Everything I do related to clocks is new to me so most things end up being done at least twice. I just hope that what I learn I don't forget for the next one. I think maybe I need to start taking notes to compensate for my poor short term memory.

    Phil:)
     
  33. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
  34. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Thanks for the input, that's an interesting point. What's the technique for vertical graining, how do you control the process to ensure that the graining is straight and vertical.

    Phil:)

    PS: I think you may have identified the problem.:D
     
  35. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    I have not done a round one yet but when I have had a dial centre to do I do it my hand.
     
  36. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Yes but what do you do by hand, vertical or circumferential?

    Phil:)
     
  37. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    sorry, dial centres are grained vertically. They were silvered on late 18th century longcase clocks before they went to flat sheet dial plates that were all over silvered, those were grained vertically too. I haven't done one of those either. However I have done dial centres by hand.
     
  38. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Hello again, sorry but I still do not understand. You say "dial centres are grained vertically". and "However I have done dial centres by hand".
    How do you do the centers by hand, vertical or circumferential?

    Phil:)
     
  39. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    Dial centres are grained vertically and I have done them by hand. IE Freehand
     
  40. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    OK I got there eventually. So you do it vertically but you control straightness and orientation by hand only.

    Phil:)
     
  41. novicetimekeeper

    novicetimekeeper Registered User
    Donor

    Jul 26, 2015
    4,821
    9
    38
    retired and on my second career
    Dorset
    Country Flag:
    Region Flag:
    yes, I'm not the only one who does that, there is a training video on here somewhere about resilvering and that guy does it by hand too on a flat sheet arched dial. a bracket clock so a bit smaller than a longcase. I have seen people use a jig to do chapter rings which are grained as a circle but I've also seen that done freehand.
     
  42. Ralph

    Ralph Registered User
    Donor Sponsor NAWCC Member

    Jan 22, 2002
    4,395
    6
    38
    Country Flag:
    I agree with Nick, the graining should be straight and true. Someone with your skill should not have any problem making up some sllding rig that your abrasive would follow.

    On the hour arbor, some regulators have the arbor with a mounted matching wheel to the great wheel, 1:1 and drive wheel to wheel... no pinion.

    Your movement looks very nice. The performance is outstanding.

    Ralph
     
  43. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Here’s a Microset plot that shows a strange transient rate change for a permanent change in ambient temperature of 4 degree C.

    At around 17:00 on the first day the heating in the room was turned off an the temperature allowed to fall. Almost immediately the rate slowed rapidly and was mirrored by a similar rapid increase in amplitude, as you might expect. What was not expected was that this slowing of the rate was produced by a falling temperature. The bob was suspended from its centre using a clamp, without the temperature compensation device being active.

    Half way thought the temperature drop the slowing of the rate and increase in amplitude stopped and then reversed with a speeding up of the rate but surprisingly without any change in amplitude. When the temperature reached its lowest value the rate continued to speed up, indicating a high degree of lag relative to the ambient temperature. The rate finally returned to slightly above the starting point, when the temperature was 4 deg C higher. During all of the temperature increase the amplitude remained constant.

    This transient change in rate resulted in an equivalent permanent drift from the true time of approximately 0.5 seconds which is not good for the precision I am aiming for, consequently I need to understand what’s going on so I can hopefully eliminate or at least mitigate the effect.

    I have observed this transient effect a number of times and below is a typical plot:

    Does anybody have any theories as to what is going on?

    View attachment 340486

    Phil:)
     
  44. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,680
    28
    0
    Calif. USA
    What is more interesting is what would take 8 or 9 hours
    to settle down? It is hard to think what might have
    that kind of time constant?
    Tinker Dwight
     
  45. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    Hi Tinker, thanks for the input. I think there are two clues: the amplitude doesn't change during the 8 hours and the rate of change of the rate over the 8 hours is constant (no tailing off as it approaches equalisation). These two clues seem to indicate that the issue is not thermal expansion/heat capacity related, but what then? I think I may tried to see if I can repeat the result!

    Phil:)
     
  46. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,680
    28
    0
    Calif. USA
    It may be harmonic resonance. In the past I've mentioned that
    things like where the crutch attaches to the pendulum can create
    resonances in the rod. These can cause the pendulum to pull slightly
    off frequency if they were close to harmonics of the pendulum.
    If the harmonics were high enough, you could have a step like action
    as the pendulum changes frequency.
    Tinker Dwight
     
  47. Phil Burman

    Phil Burman Registered User

    Mar 8, 2014
    93
    1
    6
    Suffolk England
    Country Flag:
    I'm still struggling with the temperature compensation. Contrary to my previous proclamation the pendulum didn't have zero change without any temperature compensation. It now seems that my invar rod has a coefficient of expansion 0.0000021 mm/mm/deg C which is about right for annealed invar. It seems that cold drawn gives the normally quote value of 0.0000011 m/mm/deg C. I'm about to try a 170mm aluminium compensation tube and hopefully that will nail it.

    In the meantime I've be working on the barometric compensation. I was lucky to obtain 2 of the correct type of aneroid capsules (self sprung) off ebay. I built a low cost vacuum chamber in order to calibrate them, see photo. The system uses a dial indicator for deflection measurement, the Microset atmospheric sensor for pressure and a low cost brake bleeding vacuum pump. In hindsight a good pair of lungs would probably have been sufficient.

    I have included a graph of the result for one of the capsules, the other gave almost the same result.

    View attachment 355650 .................. View attachment 355651
     
  48. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,680
    28
    0
    Calif. USA
    :)
    Tinker Dwight
     
  49. jhe.1973

    jhe.1973 Registered User
    Donor NAWCC Member

    Feb 12, 2011
    448
    2
    18
    Arizona, USA
    Country Flag:
    Region Flag:
    Hi Phil,

    Thank you so much for keeping us informed of your progress with your clock and experiments/developments.

    Your thread is a keeper for sure!

    :thumb:
     
Loading...
Similar Threads - second Regulator – Forum Date
Jewelers regulator 3 jar mercury pendulum finding center of oscilation? Clock Construction Apr 10, 2017
Astro Regulator Clock Construction Mar 29, 2016
John Wilding English Regulator Clock Clock Construction Jan 6, 2015
New walnut jeweler's pinwheel regulator case! Clock Construction Mar 15, 2014

Share This Page