This puzzled me since yesterday as I discovered how the suspension of pendulum clocks is made, that is a springy plate. I thought that: -the suspension point (i.e. the point the pendulum moves about) cannot be considered fixed, -following the above, the lenght of the pendulum varies during a swing, -the spring contributes to the oscillation (it cumulates and releases energy), i.e. not only gravity at work.. Ok that we can neglect the above yet having an extremely good approximation, but is it conceptually right? Thanks for the answers. ant

List -- we don't normally do wrist watch or pendulum topics here on time-nuts. There are many publications [1] and forums [2] for that Iovane -- but, here's a quick answer: > This puzzled me since yesterday as I discovered how the suspension of pendulum clocks is made, that is a springy plate. > I thought that: > -the suspension point (i.e. the point the pendulum moves about) cannot be considered fixed, > -following the above, the lenght of the pendulum varies during a swing, > -the spring contributes to the oscillation (it cumulates and releases energy), i.e. not only gravity at work.. > Ok that we can neglect the above yet having an extremely good approximation, but is it conceptually right? > Thanks for the answers. When you get to 0.1% or ppm or ppb levels, there is a huge difference between pendulum motion as described in physics textbooks vs. how real or precision pendulum clocks operate. Correct, it's not all gravity; the spring suspension itself has some effect on the motion of the pendulum. If this is a problem then you make design decisions about spring length, width, thickness, taper, or what metal to use. Or make the bob heavier, or choose a different operating amplitude. Or skip the spring entirely and use agate or diamond pivot suspension. All of this can be worked out with math or with experimentation. Correct, the effective length of a pendulum may vary during the swing. This is due to the geometry of the suspension, flexure in the rod, and changes in buoyancy. And even if the length were constant, the pendulum is still not isochronous, due to "circular error". Hence the preoccupation with amplitude stability as well as length stability. In general you should not neglect anything until you have modeled or measured it. Also don't confuse accuracy with stability. A pendulum could swing in a "figure 8" for all I care, but as long as it does so consistently it can be a good timekeeper. It's possible that some of the effects you describe affect accuracy more than stability. Given your curiosity about pendulum clocks, I strongly suggest you subscribe to HSN and order the digital archive. These topics are covered in extreme detail there. You will be amazed. /tvb [1] Some pendulum clock resources: AH (Antiquarian Horology), www.ahsoc.org HJ (Horological Journal), British Horological Institute, bhi.co.uk/horological-journal NAWCC (National Association of Watch & Clock Collectors), www.nawcc.org HSN (Horological Science Newsletter), NAWCC chapter 161, www.hsn161.com [2] There are also many clock and pendulum web sites, web forums, and mailing lists. Contact me off-list.