How Does The Timegrapheror Vibrograph Measure Amplitude?

[cdjswiss]

My question is in the title: How does the Timegrapher or Vibrograph measure amplitude? Can anyone provide an answer?

 

The frequency of Simple Harmonic Motion does not depend on amplitude. So to first oder everything that can be measured by the Timegrapher is independant of amplitude. Does it use a second-order effect caused by the balance spring to extract the amplitude from the signal and, if so, to to what precision?

 

Or am I missing something simple?

[Geo]

Hi Colin, have a read through this thread, there may be some clues.

http://www.watchrepairtalk.com/topic/604-do-it-yourself-timing-machine/

It might be an idea to PM Stefan and see what his take is on it, he must have some idea having built his own machine.

[cdjswiss]

Thanks Geo for the link. I shall PM Stefan - some time ago read his post.

 

It must be that the angular speed of the roller jewel is measured - this does depend on the amplitude. The speed can be calculated from the time difference between the sounds of roller jewel hitting the fork and the fork hitting the banking pin. But this will only be corrrect if the beat error is zero and the distance from the fork to the banking pin is known to the Timegrapher. Also there will be instrumental measurement errors.

 

I ask all of this because I strongly suspect inaccuracies in the measured values from my generic Timegrapher.

[Geo]

I remember you saying in the past that you thought there were issues with your machine. I hope Stefan can be of assistance.

[Rob]

Witschi has technical explanations of how the values are calculated. The amplitude is calculated from the noise of the escape-wheel tooth meeting the locking plane of the pallet-stone and the lever hitting the banking-pin.

 

"To calculate the amplitude, the time between the first pulse and the third pulse of the beat noise is measured.

 

Between these two pulses, the balance wheel rotates a certain angle. This so-called lift angle is determined by the construction of the movement and is entered as a parameter. The larger the amplitude of the balance wheel, the greater is the speed with which it goes through these lift angle and the shorter is the time it needs to traverse this angle.

 

The amplitude can therefore be calculated from the time between the first and the third pulses in the beat noise, taking into account the beat number and the lift angle."

 

[cdjswiss]

Thanks, Rob, for this link and technical information - it is exactly what I was looking for.

 

I conclude that my 'Timegrapher' is US. It always displays a beat error of 5.1ms irrespective of the separation of the two plots and the amplitude is another fiction. The timing information is reasonablly good but the rest is trash, presumably because of bad resolution of the individual pulses of the beat signal.

 

Perhaps the mike is faulty?  I shall look at the mike output on my 'scope to check that the three pulses are resolved. But almost certainly I shall be investing in a better instrument.

[cdjswiss]

I finally got up-to-date on Watch-O-Scope. This may be my salvation as I am awaiting delivery of a Greiner microphone.

 

Many thanks to Geo, Rob and Stefan.