Re: [Phys-l] how to prove relativity

[chuck britton <cvbritton@mac.com>]

'Spoken' like a true horologist ! ! ! !

I sit aMAZed by the resources that are shared here.

At 9:29 PM -0700 6/2/10, Bernard Cleyet wrote:
> Still no way.
>
> Relativity
>
> Satellite clocks are slowed by their orbital 
> speed but sped up by their distance out of the 
> Earth's gravitational well.
> A number of sources of error exist due to 
> relativistic effects[82] that would render the 
> system useless if uncorrected. Three 
> relativistic effects are the time dilation, 
> gravitational frequency shift, and eccentricity 
> effects. For example, the relativistic time 
> slowing due to the speed of the satellite of 
> about 1 part in 1010, the gravitational time 
> dilation that makes a satellite run about 5 
> parts in 1010 faster than an Earth based clock, 
> and the Sagnac effect due to rotation relative 
> to receivers on Earth. These topics are examined 
> below, one at a time.
>
>
> clock much better than ~ one part in 10^10  required
>
> http://en.wikipedia.org/wiki/Global_Positioning_System#Relativity
>
>
> and:
>
>
>
> Temperature compensated crystal oscillators 
> typically employ a thermistor network to 
> generate a correction voltage which reduces the 
> frequency variation over temperature. The 
> correction voltage is usually applied to a 
> varactor diode in the crystal circuit such that 
> the crystal frequency may be varied by a small 
> amount. TCXO stability can approach 0.1 PPM but 
> several problems must be addressed. A TCXO that 
> resides at one temperature extreme for an 
> extended period of time may exhibit a frequency 
> shift when returned to normal room temperature. 
> Usually this hysterisis (sic) or "retrace" error 
> is temporary but a seemingly permanent offset is 
> common. Retrace errors are usually less than 
> about 0.1 PPM but can be much higher, especially 
> if the mechanical tuning device (trimmer 
> capacitor or potentiometer) is shifting. This 
> hysterisis (sic) makes the manufacture of TCXOs 
> with specifications approaching 0.1 PPM quite 
> difficult. The high precision crystals found in 
> oven oscillators exhibit less retrace but they 
> are unsuitable for use in TCXOs because they 
> often exhibit activity dips at temperatures 
> below the designed oven temperature and SC-cuts 
> and high overtone types cannot be tuned by a 
> sufficient amount to compensate for the 
> frequency excursion.
>
> vs.  one part in 10^7
>
> http://www.wenzel.com/documents/tcxo.html
>
> bc
>
> p.s.  Here's a TCXO GPS corrected for long term stability:
>
>
> http://www.bmumford.com/mset/GPSref.html
>
> p.p.s.  Cesium clock necessary:
>
> http://tycho.usno.navy.mil/cesium.html
>
> note claim that the Cs clock:  "... is the most 
> accurate realization of a unit that mankind has 
> yet achieved. A cesium clock operates by 
> exposing cesium atoms to microwaves until t
>
> I suspect Fe 57 Mössbauer 14 KeV transition is more stable.
>
> nope only 1/ 10^12  for Fe-57
http://www.rsc.org/membership/networking/interestgroups/mossbauerspect/intropart1.asp