ATO Question about the dampening / eddy current tube size for an ATO clock

Discussion in 'Electric Horology' started by mopydick, Aug 18, 2016.

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

    mopydick Registered User

    Jul 12, 2016
    91
    0
    6
    Country Flag:
    I'm working on my first ATO clock. It suffers from the common problem on losing amplitude over time until it stops. I've removed the movement to narrow the issue (is it the load of the movement or lack of power from the electro-mechanical drive). There was no change in behavior (still lost amplitude over time). The dampening / eddy current tube (I don't know its proper name) looked a bit suspicious to me. Its a section of standard copper pipe about 3/4" long that someone had clearly fabricated at home. I removed this and the pendulum amplitude seemed very healthy. For the time being, I have some packing material on the end stop to keep it from knocking.

    So I need to fabricate a new dampening tube of the correct size and material.

    Would anybody know the approximate length that I should make the tube. I'm tempted to make it out of brass instead of copper since the rest of the clock is brass. Could someone tell me what the original tube material was?

    Thank you and you guys have a really great forum here. It's my first visit but I'm sure I'm going to be spending more time hanging around,

    Peter



    pix225.jpg pix226.jpg
     
  2. eskmill

    eskmill Registered User
    NAWCC Fellow NAWCC Member

    Aug 24, 2000
    7,135
    27
    0
    Region Flag:
    #2 eskmill, Aug 18, 2016
    Last edited: Aug 18, 2016
    Welcome to the Message Board Mopydick

    The "damper" should or would be similar to the bracket that holds the drive coil. I really doubt that a lengthy copper or brass tube is necessary for the nominal ATO drive voltage of a single 1.5 Volt cell.

    Try this: make a one-turn damper coil from a length of #10 or two turns of #12 heavy copper wire as a substitute for the foam packing you've used to eliminate "knocking."

    Reply with your results.
     
  3. Hans Vrolijk

    Hans Vrolijk Registered User
    NAWCC Member

    Jan 7, 2005
    174
    0
    0
    retired family doctor
    Rotterdam, The Netherlands
    Country Flag:
    Region Flag:
    Here's a picture of the copper tube, 13 mm length, in one of my Ato's.
    DSC01197.jpg
     
  4. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    63
    0
    Calif. USA
    Hans, he'll need the diameter as well.
    I would think one can just shorten the one you have with
    a file until is has reasonable dampening.
    Tinker Dwight
     
  5. mopydick

    mopydick Registered User

    Jul 12, 2016
    91
    0
    6
    Country Flag:
    Well, one thing leads to another. I made up several different lengths of copper tubes (5mm, 10mm, 15mm, 20mm), but before I started experimenting, I thought it would be a good idea to instrument the voltage across the electro-magnet coil with the white packing foam still installed as the dampener. A oscilloscope screenshot is shown below.

    Some observations and assumptions (please correct me if I'm wrong)
    1) The points are in the open state about 40% of the time, and close (current flowing) 60% of the time. (Is this normal?)

    2) The 'hump' on the 1.5v part of the waveform (circuit closed), is a result of the BEMF of the magnet has it initially starts to pass into the electromagnet (EM), increasing to its high point when the permanent magnet (PM) on the pendulum reaches its 'turn around point', and then the back side of the 'hump' is when the pendulum is getting its impulse as the PM swings back out of the EM. Finally, the 1.5v drops to 0v as the points open.

    3) The chatter on the voltage after the points have opened is probably due to the points needed some cleaning.​


    Other than the chatter, I thought everything looked pretty good electrically.

    But then I started to notice some truncated 'on' periods (2nd & 3rd scope shot).

    On the 2nd scope pix, I have the cursors set to show that a normal on period is about 276ms.

    And on the 3rd scope pix, I have the cursors still set at 276ms, but you can see that the start period didn't start when it was expected to do so. In fact, there is a little blip of voltage at the expected start time, which to me, indicates that the points tried to complete the circuit.

    So, I'm now going to turn my attention to cleaning the points before I go any further with the eddy current dampening ideas.

    Thanks for listening and input,

    Peter

    pix225.jpg

    pix226.jpg pix227.jpg
     
  6. eskmill

    eskmill Registered User
    NAWCC Fellow NAWCC Member

    Aug 24, 2000
    7,135
    27
    0
    Region Flag:
    Thanks for the 'scope analysis. In my opinion, I believe the contact closure and resultant current flow to the drive coil is grossly long. I would think the contact should be very brief, else there would be a tendency to "grasp" and control the motion of the pendulum.

    In my view, the concept is to provide only a restoring force to the pendulum and to allow gravity to control the motion of the pendulum magnet rod. To the contrary, your analysis tends to behave much like a "motor" with the electromagnetic coil driving a free running armature.

    The timing of the contact closure is most critical I believe.

    JMO, Les
     
  7. praezis

    praezis Registered User

    Feb 11, 2008
    413
    15
    18
    Germany
    Country Flag:
    Region Flag:
    Hello Peter,

    in your pictures the drive pulse is negative (going to zero) not positive. So in your 3rd picture you have one too long driving pulse (just in front of the marked pulse).

    I do not know this type of Ato, but I agree with Les, the driving pulses seem to be too long. My Junghans Ato has a period of 1 second with driving pulse 100ms.

    I only found an oscillogram of my Brillie for comparison, but it is the same system (drive positive to 1.5V):

    Brillie_osz2.jpg

    Frank
     
  8. mopydick

    mopydick Registered User

    Jul 12, 2016
    91
    0
    6
    Country Flag:
    Hello All,

    I see what you guys are saying about how I was mis-interpreting my waveform. For some reason I was thinking that the BEMF voltage was the drive period (and I hate to admit that I've worked with PM motors a good deal in my life).

    I'm about to disassemble the pendulum section to inspect everything and give the points a cleaning. Clearly I need to be thinking beyond just the dirty points as the issue, since the overall timing of pulse delivery is suspiciously long.

    Out of curiosity, I took a scope shot of the coil voltage with it swinging freely (not being driven electrically driven). The two differences that jump out are the difference in frequency (1Hz vs 2Hz) and the coil voltage amplitude. I've wondered about the strength of my permanent magnet. Just how permanent are these, being 100+ years old? Is re-magnetizing the PM another task I should add to my list, or is this(low magnetism) a unlikely source of trouble?

    Thank you everyone for your input; your comments are keeping me on track.

    Peter


    pix228.jpg
     
  9. eskmill

    eskmill Registered User
    NAWCC Fellow NAWCC Member

    Aug 24, 2000
    7,135
    27
    0
    Region Flag:
    mopydick has a concern about the strength or retentivity of the pendulum curved bar magnet. I don't think it's an issue based on his earlier posts in which he stated that the end of the pendulum bar required a cushion.

    I have stated in an earlier reply, "that In my view, the (ATO) concept is to provide only a restoring (feedback) force to the pendulum and to allow gravity to control the motion of the pendulum magnet rod." A restoring or feedback impulse is minimal and compensates for losses caused by external factors such as friction.

    I like to think of inventor Hatot, Lavet and Moulons, along with Ferey's contributions to electrical timekeeping more like the principles of electrical or electronic oscillators wherein the most efficient and accurate oscillators are those that require minimal feedback to sustain oscillation.

    The pendulum as an oscillator too requires feedback to sustain oscillation from gravitational force.

    The feedback impulse, whether mechanical or electrical must occur at the most effective instant when gravity is driving the pendulum at its maximum rate and not a fraction of a degree later. A late feedback or "restoring" impulse counter-acts the force of gravity and retards.

    Thus the instant of contact closure to the battery-drive coil must occur at a point in the pendulum motion that Gravitational force is imposing its maximum force in the POSITIVE direction. In a purely mechanical pendulum, this point in the pendulum travel is immediately before the center of pendulum motion. A restoring or feedback impulse delivered to the pendulum oscillator after the pendulum crosses the middle of its travel opposes the force of gravity.

    In the ATO battery clock, the instant of contact closure must occur while the pendulum is moving forward at its maximum rate.

    JMO Les
     
  10. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    63
    0
    Calif. USA
    Looking at the different waveforms, I'd say your clock is driving the
    pendulum for too much of the cycle.
    You need to adjust the amount of time the points are closed to look
    more like Franks waveform.
    You may want to ignore inversion of the wave form as that may
    just be a hookup issue.
    Tinker Dwight
     
  11. mopydick

    mopydick Registered User

    Jul 12, 2016
    91
    0
    6
    Country Flag:
    Didn't have any time to tinker with this clock last week but I'm at it again.

    I separated the pendulum assembly from its support structure to get better access to the points for inspection (photos below). The points certainly look like they could use a cleaning. Is a simple light polish with some steel wool sufficient? Or possibly too aggressive? Or is it standard practice to replace them?

    Also, I got to thinking about the electrical pathway, and how it passes from the points on the pendulum, through the shaft between the yoke, and back to the frame ground of the whole structure. The shaft within that yoke has an axial preload on one end. Originally, I thought that was there for mechanical purposes. But now I'm thinking its yet another intended point of electrical connection. The preload spring's job is to keep the circuit closed while the shaft rocks back and forth. If my speculation is correct, I assume that I should give the contact point between the preload spring and the end of the shaft a quick cleaning/polishing.





    1.jpg 2.jpg 3.jpg
     
  12. mopydick

    mopydick Registered User

    Jul 12, 2016
    91
    0
    6
    Country Flag:
    Pendulum impulse point


    I'm going to try to tread lightly here, as I'm the new guy, and I suspect this topic has been hashed over many times in the past.

    But, when I reach back into my physics background, and remember using energy based analysis of mechanical systems, I don't see why an ideal mechanical impulse needs to be at any specific point in the arc of the pendulum, as long as its in the direction of travel of the bob. The power imparted, whether at the beginning of the swing, mid point, or near the end, will still be the same impulse of power. And the system's net energy will remain constant, provided that the impulse power equals the power losses due to friction, windage, etc.

    My statement applies only to the idealized, fictional world that we study in school. Obviously, in the context of discussing an ATO clock, the timing of the electrical impulse is going to need to include several other factors like the shape of the force curve the electromagnet delivers when electrically pulsed, and the shape of the opposing PM field as it approaches and/or departs the electromagnets fields. I would have guessed the appropriate place to pulse the electromagnet would be near the end of the PM so that the EM field 'sees' the greatest PM field and generates maximum force.

    Again, I'm not trying to start a fight, just a discussion so that I can gain some insight on how these ATO's really work.

    Thank you,

    Peter
     
  13. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    63
    0
    Calif. USA
    Re: Pendulum impulse point

    This would be true if it were an exact true instantaneous impulse
    but in real life it happens over some time and is not always exactly
    the same amount.

    So, now we have to look at how to drive the pendulum with a pulse
    that might have 1-2% variation and still have the pendulum keep good time.
    If one assumes that a free swinging pendulum is mostly isochronal
    ( not counting circular error ), one finds that the best place to change the
    velocity of a pendulum is as it crosses the center of swing and not at some
    arbitrary point.

    If one were able to change the restoring force, one would see that the ends
    of the swing would be best. It is difficult to change g ( gravity ) but some
    attempts have been done with modified suspension springs, on pendulum clocks.

    It is a problem of disturbing the pendulum with an inaccurate impulse and still
    get the most consistent rate from the pendulum.

    Considering the angle of swing of the ATO type clocks they are remarkably
    fare time keepers. I suspect that without the eddy current dampeners they
    would not do as well.
    Tinker Dwight
     
  14. eskmill

    eskmill Registered User
    NAWCC Fellow NAWCC Member

    Aug 24, 2000
    7,135
    27
    0
    Region Flag:
    Re: Pendulum impulse point

    One important fact to keep in mind when discussing the ATO and Bulle battery clocks is that both are "WEAK CURRENT" battery powered timepieces.

    The "feedback" or "restoring force" is small when compared to Gravity.
     
  15. mopydick

    mopydick Registered User

    Jul 12, 2016
    91
    0
    6
    Country Flag:
    Some more progress ... after enough disassembly to get clear access to the contact pins, I was able to do a light cleaning of the pins. Reassembled and found that it timing of the pulse just wasn't were it ought to be (other discussions aside, it would seem that the pulse should happen near the 'turn around' point of the pendulum where the PM field is the strongest. With some very careful tweaking of the contact pins (and it doesn't take much), I was able to get the timing of the pulse where I wanted (more or less). Its still too wide, but I grow weary of making an infinitesimally small tweak to the pin that either does nothing or moves the timing too much.

    The pendulum ran all night with an incredibly healthy amplitude - now I can see why I need the eddy current dampener!

    Next up, a final thorough cleaning of the pendulum section. I've been in and out of there too many times and have left fingerprints everywhere. And then a slow reassembly of the rest of the clock (I believe in 'baby steps' too see what affects what).

    pix01.jpg

    [video]https://youtu.be/kz7axB3WgOQ[/video]
     
  16. mopydick

    mopydick Registered User

    Jul 12, 2016
    91
    0
    6
    Country Flag:
    Now things are starting to come together. I might even get back to the original topic of Eddy current dampeners!

    I gave the pendulum assembly a cleaning in the ultra-sonic cleaner, which not only made things look bright and new, but it really cleaned up whatever junk was on the contact pins. The 'on/off chatter' of the electrical signal is completely gone now! I know, the timing and the period may not be spot on, but after playing around with this clock for 2 weeks, its nice to see some improvement!

    My last (most recent?) conundrum is whether I should have lubricated the pivots on the bar the contact points are attached to (see pixs below). I've pretty much convinced myself that the pivots themselves are there for their obvious mechanical purposes AND the one end of the pivot is special in the sense that it has a light leaf spring pressing up against it to maintain continuous electrical continuity. It even appears to be made of a different material than the rest of the pivot, at least it appears to be a different color in my pictures.

    I was going to lubricate the pivot bearings with a light oil, but then I decided not to, out of fear that the oil could possibly creep to the tip of pinion and interrupt the electrical continuity. I know, I know that a properly oiled pinion shouldn't weep/creep oil but I'm still fearful of this, being a bit heavy handed sometimes. And the local clock guys around here tend to shy away from the use of any oil whatsoever on low load pivots (like an escapement).

    Next item of interest for me, is added the Eddy current dampener and seeing how it effects the amplitude and if it changes the frequency of the pendulum. From my scope shots, you can see that I'm running a bit fast at the moment ...

    01.jpg 02.jpg 03.jpg 04.jpg
     
  17. sophiebear0_0

    sophiebear0_0 Registered User

    Nov 5, 2012
    59
    6
    8
    Folks

    Some interesting comments and sharing of experience as usual. All very helpful.

    I have worked on a number of different ATO models. So I thought it might be helpful to share some of my experience:

    1) I have checked in John Locke's book and he claims that the spring contact is primarily present to add friction so that the contact block needs to be pushed side-to-side rather than be free to swing under gravity.

    2) My experience is that the spring contact does form part of the electrical conducting path and should be burnished/cleaned. Thereare a range of designs and some have a metal "button" fixed to the steel spring.

    3) John Locke recommends using 0000 steel wool to clean contacts. I prefer to use electrical contact fluid and a small rubber brush (inter-dental cleaners)

    4) The build-up of oxide layers due to sparking occurs and the contacts need to be cleaned on a regular basis. The rocker contacts are designed to be self-wiping but my experience is that this is not 100% effective. It is interesting to note that for the ATO models that have non-swivel contacts (horizontally mounted) included a 5k resistor in parallel with the coil to suppress sparking. This seems to work quite well - albeit with a slight extra drain on the battery.

    5) I'm not sure that the length of the damping coil is critical - though I think it is critical to include a damping coil. Some of the ATO designs have just a ring which is around 1-2mm wide, others are much longer tubes (> 10mm)

    6) I usually run-in my ATO's with a slightly higher voltage (3 v max) to help bed in the contacts and the movement (I am not sure the purists would agree with this approach - but i have never had a coil failure dues to extra current). This approach can also sometimes help cure a low oscillation amplitude due to poor contact wiping. The advantage is that this can be done without removing the movement - but off course is no real substitute for carrying out a proper electrical cleaning.

    7) My experience is that not all the coil resistances are the same ? Some models are around 1200 Ohms whereas others appear to be close to 3000 Ohms. That said, I have always been able to get the clocks to run correctly on 1.5 volts.

    8) My experience is that the ATO magnets are good at retaining their magnetic properties. From memory, I think I have only ever re-magnetised one ATO magnet - and that may not have been really necessary. Maybe the simple N-S magnet that you find in the ATO's is more able to retain its field better than the S-N-S filed used in most Bulles (just guessing ?)

    People might want to know that the John Locke book is available electronically at low cost on a well known auction website. I found it quite a useful reference - but sadly there is nothing comprabale to Pete Smith's great documentation on Bulle clocks - or at least nothing I am aware of ?

    Best regards,

    Peter
     
  18. mopydick

    mopydick Registered User

    Jul 12, 2016
    91
    0
    6
    Country Flag:
    Home stretch ... ?

    It ran fine naked on the bench today. I just finished putting it back in the case. So far, so good.

    One last question, and it seems silly. But what the heck is the propeller like thing on the back of the brass case? On the clock I have here, it doesn't connect to anything internally.


    Thank you everyone for chiming in. I've learned a lot on this cycle,

    Peter

    PS. Sophiebear0_0, thank you for your comments. They've reinforced much of what I've learned. And just as a reference point, my coil resistance measured 3300 ohms. And now that I've worked on one of these clocks, I should buy one of the books you've mentioned :)



    pix05.jpg 05.jpg
     
  19. sophiebear0_0

    sophiebear0_0 Registered User

    Nov 5, 2012
    59
    6
    8
    Glad you have got everything sorted.

    The propeller is simply to hold the original battery in place. The original battery is a rectangular block with 2 terminals at the top which mate with the terminals at the top of the base plate. ATO battery.JPG


    One word of warning - if you decide to make a replacement battery holder, be aware that the a steel casing can create havoc with the magnetic field. In fact this can be a problem if you mount steel cased (modern) batteries too close to the magnet.

    Some of the ATO clock actually have small supplementary magnets mounted on the exterior of the case. In theory these can be used to fine tune the clock regulation. My experience is that these are not very effective - I suspect that the impact is reduced because of the modern batteries with steel casings ?

    Regards,

    Peter
     
  20. Tinker Dwight

    Tinker Dwight Registered User

    Oct 11, 2010
    13,666
    63
    0
    Calif. USA
    I can see the line with the word voltage but the rest of the line is hard to
    make out?
    Tinker Dwight
     

Share This Page