Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Measuring Time: A Revolutionary Method

  1. Jun 13, 2004 #1
    Three identical flat disks whose thickness T<<<D , the diameter, are arranged with their axes of rotation mutually perpendicular, are calibrated in a stationary frame such that the three total number of revolution counts are exactly the same. Each revolution of each disk is counted by a separate high memory content Chinese Abacus.

    To each axle of rotation is a feedback motor control used to correct for friction losses only, as measured by measuring revolution count differences in each disk. The control systems are independent of each other.

    Or, replace the system above with frictionless spinning disks, at a slightly greater cost. On sale, this week only! "Friction < 0" SuperGrease.

    Accelerate the frame containing the spinning disks along an axis parallel to one of the axes of rotation to some constant velocity. Keep a running count of the total revolutions, A, B and C, of each disk and simultaneously measure the |error|, differences in current total number of revolutions, A, B and C, of each disk as each error is expressed as:

    a. |A - B| ?=

    b. |B - C| ?=

    c. |A - C| ?=

    What are each "?=" in terms of >0, <0, or =0?

    You select the axis parallel to the direction of motion.


    Confirming Rumor: Einstein replaced the mechansim of mechanical clocks with atomic clocks re SR. Can anyone verify this rumor? If yes, is AE's reasoning ever explained by him, or anyone else? If yes, what was the stated reasoning?
  2. jcsd
  3. Jun 13, 2004 #2


    User Avatar

    Staff: Mentor

    Atomic clocks are the standard these days for a pretty simple reason: they are the most stable and precise clocks available.
  4. Jun 13, 2004 #3
    Another reason: Mechanical clocks may not give the atomic clock answer. You didn't answer all the questions in the post. Is why your answer is so spartanly offered?
  5. Jun 13, 2004 #4


    User Avatar

    Staff: Mentor

    That's the same reason I gave. Mechanical clocks don't give the atomic clock answer because they are incapable of the level of precision required.
    The rest of your post is meaningless gibberish.
  6. Jun 14, 2004 #5
    So mechanical clocks give different answers than atomic clocks?

    Do you have any references that unambiguously supports the 'precision' claim?

    Saying it doesn't cut it anymore. You make a statement, prove it or stay off the thread, understand?

    And finally, you don't know the answers to the specific questions, do you?
  7. Jun 14, 2004 #6
    seems to me rotating gears could get VERY precise...

    Code (Text):

    |----------Big Gear----------||--Little Gear--|
                            |----------Big Gear----------||--Little Gear--|
                                                    |----------Big Gear----------|
    just keep stepping up the gears etc. get to where one rotation of the first gear yields 35000 rotations of the final gear... course i'm sure the power needed to drive the thing would be substantial :D
    Last edited: Jun 14, 2004
  8. Jun 14, 2004 #7
    they need to add fixed space text preformatting :O
  9. Jun 14, 2004 #8
    I'm not sure what you mean. As I originally proposed, the three spinning disks have mutually different 90 degree axis variation. As some have suggested, motion transverse to the frame direction is invariant under SR (see the recent shortened space ship thread), therefore the disk spinning with the flat side axis of rotation in the direction of motion should be 'relatively' invariant under motion re SR. The other two disks all have a sin(theta) addition/subtraction of revolutions to the velocity of the frame. However you do it the frame with parallel axis to the direction of motion should read a different total number of revolutions wrt the other two rotating disks. The decay in revolutions, if any, could be compared to an atomic clock.

    Even with a predicted (by russ_waters) accuracy decline, the rates of change of all three can still be monitored for variances.

    Explain your 'geared' configuration, por favor.. Why so confident in a very accurate result?

    Also, if frictionless axes are unacceptable then simply measure the friction induced decay rate of all three disks. The Swiss have passenger busses powered by large weight rotating masses. Put some of these on board a space ship and count revolutions.

    "Basic SR', as Tom_Wattson would say.

    russ_watters says the reason to go to atomic clocks presently is because of resolution differences,AKA accuracy betterment. Myself, I think it is because of revolution differences in mechanical v electronic (or atomic) clocks. Anyway, russ_watters avoided the question of why Einstein suddenly adopted atomic clocks in his theoretical development, while discarding, with no observed comment, the mechanical models.
    Vertlich, sehr fremde!
  10. Jun 14, 2004 #9


    User Avatar

    Staff: Mentor

    No. Mechanical clocks give no answers at all to these questions.
    Well you tell me - have you ever seen a mechanical clock with a readout in nanoseconds? Since a mechanical clock is read manually, its precision depends on the reflexes of the person using it. And the accuracy (you know the difference between precision and accuracy, right?) of the best mechanical clocks is on the order of a few seconds per month.
    Do you have any idea how funny that sounds? You're the one here making the outrageous claims and providing no support whatsoever for them.
    Einstein likely assumed (as did I) that anyone posessing enough knowledge of physics to delve into Relativity would know that mechanical clocks aren't capable of the precision/accuracy required to be useful in science. I won't make that mistake again.
    The flaw in your device is quite simple: it isn't capable of regulating its own speed. Its missing the component that makes a clock a clock (a spring, a pendulum, a chrystal, cesium gas, etc). Unless of course the "controller" has a clock in it...
    Last edited: Jun 14, 2004
  11. Jun 14, 2004 #10
    Russ_watters, please take another look. we can arbitrarily assign the revolutions of the discs as basic time machines, clocks. Assuming we achieve a high revolution rate, the disks are massive enough to stay running, even with a minimized friction, the disks that are spinning in the direction of motion, two of the disks, can be monitored to determine if the decay rate differences with the disk whose axle of rotation is parallel to the direction of motion (the spininning is 90 degrees to the motion of the frame),, can be measured by differences in the total number of revolutions (or if you insist rpm, a classic clock rate). The disks are clocks that haven't been calibrated in standard seconds, minutes and hours.

    Basic question: Your point of resolution aside, do you see that the arrangement as descibed is theoretically able to detect rotation decay differences depending on the direction of motion of the particles in the respective disks? Given enough time, sufficiently high velocity, sensitivity to distinguish friction losses, ought not the disks spinning in the direction of motion suffer a more rapid decay of total revolutions (measured difference in total number of revolutions) due to SR theoretical considerations, than those disks spinning 90 degrees from the frame motion?

    Theoretically speaking, Russ_watters, practical considerations aside?
  12. Jun 14, 2004 #11
    as far as gears go it's simple step up or step down based on the number of cogs.

    supposing you have a gear with 30 cogs turning a gear of 10 cogs. the 30 will turn the 10 cog gear 3x times in one revolution. attach the 10 cog to another 30 cog and use THAT one to drive another 10 cog and you've compounded the revolutions, 1 rotation of the first cog leads to 9 rotations of the final cog.

    when used for accuracy you would spin the first cog at a set rate, say 1 rotation per second but as far as PARTS of a second you could only divine 30 subsections on that wheel by counting how many cogs have passed in that example.

    but if you count cogs on subsequent gears in the series you can get very accurate because of the compounded rotations.

    in this example, if we counted 3 rotations and 13 cogs in the final wheel in one second, we calculate that the first wheel turned 0.381481 revolutions
  13. Jun 14, 2004 #12
    Why should we expect this according to SR? Have you done the calculations...?

    Oh, sorry geistkiesel, I forgot you don't like math.

  14. Jun 14, 2004 #13


    User Avatar

    Staff: Mentor

    What you are saying is if you spin them really fast, they will take a long time to slow down. They are not self-regulating.

    So they have an arbitrary rate that is constantly changing. Like I said - that makes them utterly useless as clocks.
    Theoretically, they aren't clocks. Practically, they aren't clocks. Either way, they aren't clocks and are utterly useless for scientific investigation.

    You can only measure the decay rate of their rotation if you have something to measure it against - a clock, for example.
    Now this is a separate issue: you're asking if a mechanical clock could be made accurate enough to show an SR effect (not saying that what you are suggesting actually is an SR effect) if it exists, would it? The answer is of course, yes.

    I guess you are trying to make a complicated sounding clock-in-a-centrifuge type thought experiment here. In such a case, there are both SR and GR implications.

    edit: ehh, maybe not: just looks like you are talking about linear acceleration of 3 wheels with different orientations. There are no SR implications of what is happening between the wheels.
    Therein lies the problem: how do you know its 1r/s unless you measure it with a clock?
    Last edited: Jun 14, 2004
  15. Jun 14, 2004 #14
    point taken, we can't very well say rotations are constant and consistant until we measure them up against something we know/assume to be constant, such as light speed or decay rates, or waveforms, whatever...


    but the accuracy is still there once calibrated...
  16. Jun 14, 2004 #15
    I urge you to look at Grounded in "New Theory of Light" Posty # 1. Very refreshing SR heresy.
  17. Jun 15, 2004 #16


    User Avatar

    Staff: Mentor

    Is it? I don't see any way it could self-regulate unless the "controller" has a clock in it.
  18. Jun 15, 2004 #17

    i have no idea what i'm talking about either :D

    good call
  19. Jun 15, 2004 #18


    User Avatar

    Staff: Mentor

    But at least you are thinking about it - you get cool points for that.
  20. Jun 16, 2004 #19
    I gotcha. A kind of logrythm or exponential counter, or an expanding data resolution machine without corrupting data.
  21. Jun 16, 2004 #20
    yeh but as russ pointed out, all that precision means nothing unless you have something guaranteed "constant" to weigh it against.

    just like RPMs don't mean anything without something to fill in the "minutes" part of the phrase

    oh well, back to the drawing board :D
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook