1 BNM-SYRTE Systèmes de Référence Temps-Espace Team implied in Frequency measurements André Clairon Emeric De Clercq Sébastien Bize Franck Pereira Do Santos Harold Marion Philippe Laurent Michel Abgrall Ivan Maksimovic Jan Grünert Peter Rosenbusch Céline Vian Contributions of BNM-SYRTE : Contributions of BIPM : Team implied in Time measurements Joseph Achkar Pierre Uhrich David Valat François Taris Monique Prodhomme Ishan Ibntaieb Philippe Merck Pascal Blondé Jean-Yves Richard Me Félicitas Arias Gérard Petit Peter Wolf Wlodzimierz Lewandowski Zhiheng Jiang CCTF Working Group onTAI 31 mars 2004 Jean-Yves Richard
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
BNM-SYRTESystèmes de Référence Temps-Espace
Team implied in Frequency measurementsAndré ClaironEmeric De ClercqSébastien BizeFranck Pereira Do SantosHarold MarionPhilippe LaurentMichel AbgrallIvan MaksimovicJan Grünert
Peter RosenbuschCéline Vian
Contributions of BNM-SYRTE : Contributions of BIPM :Team implied in Time measurementsJoseph AchkarPierre UhrichDavid ValatFrançois TarisMonique ProdhommeIshan IbntaiebPhilippe Merck
Pascal BlondéJean-Yves Richard
Me Félicitas AriasGérard PetitPeter WolfWlodzimierz LewandowskiZhiheng Jiang
CCTF Working Group onTAI 31 mars 2004 Jean-Yves Richard
BNM-SYRTE CCTF WG TAI 2004 2
• Presentation of the Primary Standards at BNM-SYRTE• Results• Uncertainty budget on systematic effects• Accuracy budget and uncertainties of PFS at BNM-
SYRTE• Evaluation of Collision effects • Mean frequency• Statistic uncertainty• Stability comparison FO2 - FOM• Uncertainty due to the dead times• Frequency Comparison F_EAL – F_PFS• Conclusion
3
Optical pumped caesiumbeam clock
4
53015 53020 53025 53030 53035 53040 53045
6,76E-013
6,78E-013
6,80E-013
6,82E-013
6,84E-013
6,86E-013
6,88E-013
6,90E-013
6,92E-013
y(M
aser
805
- FO
2)
MJD
y(Maser805 - FO2)
Date of measurementsIn MJD UTC unit
Uncertaintiestype A & type B of y(H-FO)(t)
NormalizedFrequencydifferencey(H-FO)(t)
Duration of integration
Bu Au maser/linku
0,50,20,8+682,5753014 – 53044
Y(Maser – FO2)Period MJD(30 days)
Results of calibration. (scale is in 1 x 10-15).
Systematic Uncertainty Statistical uncertainty
Uncertainty on theLink between
Maser & atomic Clock+ Uncertainty
due to the dead time
Mean value of y on the entire period of integration
Measurements of FO2 relative frequency fluctuations from 9 January until 9 February 2004
Statisticaluncertaintiesof y(H-FO)(t)
withoutCollisional
effects
Uncertaintieson the Collisional &Cavity pulling shift
= σAi
+ σStati
2σCollision
i
2= 6825,76 10-16y_meanMaser_FO2 by Weighted least squares with
8
With weighted least square the frequency mean for FO2 is assorted by Statistic uncertainty computed by the propagation of uncertainty from the variance covariance matrix given by least square fit :
= ( )uc yi2
+ + ( )u a12 xi
2 ( )u a22 2. xi ( )u a1 a2
Where xi is set to the middle date of the whole period :
= xi − 12Te
12Ts
= 1,63 10-16
For the period MJD53014 to 53044 it gives :
uA
Middle Date ofend measurement
Middle Date ofstart measurement
= yfit + a1 a2 xWeighted least squares linear regression gives :
9
1000 10000 100000
1E-15
1E-14
sigm
a_y(
tau)
Time (second)
Allan Deviation FO2 - FOM
2,3E-13 / sqrt(tau)
2,718E-16
Allan deviation of the difference between synchronized FO2 & FOM
ττσ
1310 x 322,2)(−
≈y
Stability of FO2 or FOM
at 2,37 days
1610 x 5,2)10( −≈= dy τσ
Stability behavior :White Noise of Frequency up to 1 day
For 10 days
162_ 10 x 6,1)30( −== du FOA τstatistical uncertainty FO2 alone :
10
Dead times of measurements on y(MaserH805 – Fontaine FO2)during the period MJD 53014 to 53044
phase data x(Maser805-Maser816) linear drift removed
MJD
x k(Mas
er80
5-M
aser
816)
/ns
The instability of the maser during each dead times gives an estimate of the uncertainty related to each idle period in themeasurement.
The knowledge of the instability of the maser is carried out starting from measurements of phase variations every hour,compared to the second maser of the laboratory.
After having withdrawn the linear slope of regression obtained by least squares, one plots the curve of the phase variationsbetween masers. One uses the Time Allan Deviation TVAR to estimate the stability of x(H805 - H816)
(suite 1)
x(Maser805 - Maser 816) from 9 January MJD 53014 until 25 February 2004
MJD 53060 slope removedStability Temporal of the phase variations between Maser 805
and Maser 816 of January 9 MJD 53014 up to February 25 2004 MJD 53060
12
For dead times of duration more than 3600s we findthe uncertainty from values of the TVARfunction.For dead time of duration less than 3600s and morethan 1800 s, we set the uncertainty with valuecorresponding to 3600s :
( )σx τ
= ( )σx tau3600s 0.10552 10-9
The uncertainty of frequency deviation is obtainedby quadratic sum of each TVAR variation dividedby the whole period of measurements of
=
second
∑ = i 1
33
( )σx ( )τm i2
T
T = 2 652 839 s (30,70416 days)
= σdeadTime 0.43 10 -15
For dead times of duration less than 1800s,we consider that the uncertainty is negligible.