Multiple phase cal tones in experiment c1013 ($01JUN25XA)
Introduction
Experiment c1013 ($01JUN25XA) was observed with channel frequency band
8 MHz channel. Correlator extracted phase cal phases at frequencies 10KHz
and 4010 KHz at X-band ant 10KHz, 6010 KHz at S-band. It was fringed two times,
with using phase cal at 10KHz and with using phase cal phase at 4010 KHz for
X-band and 6010KHz for S-band. (Exception: 6010 KHz tone at GILCREEK was used
for both cases). Two databases were created, $01JUN25XA with 10KHz phase cal
and $01JUN25X1 with 4010/6010 KHz.
Comparison of solutions
Two solutions were made, one with 10KHz phase calibration, another with
4010/6010 KHz phase calibration. Delay transfer technique was used for
comparision of these two solutions in order to be sure that solution setup,
weights, suppression flags etc were identical in both solutions.
Solution 1: with 10 KHz phase cal frequency
Delay 1371 37 ps
Baseline Statistics
Baseline # W.Obs W.RMS Del N.R.D.
used/recov ps
GILCREEK-HARTRAO 14/ 14 78 1.15
GILCREEK-SESHAN25 127/ 131 47 .86
GILCREEK-TSUKUB32 366/ 383 30 .86
GILCREEK-WESTFORD 216/ 225 34 .83
HARTRAO -SESHAN25 52/ 55 47 .75
HARTRAO -TSUKUB32 53/ 53 44 .90
HARTRAO -WESTFORD 52/ 53 31 .87
SESHAN25-TSUKUB32 316/ 331 42 .88
SESHAN25-WESTFORD 39/ 40 54 .94
TSUKUB32-WESTFORD 136/ 141 45 1.01
Corrected Reduced Chi-Square .9143
Full listing of the solution with 10KHz phase
cal tones.
Solution 2: with 4010/6010 KHz phase cal frequency
It has exacatly the same parameterization, weights, suppressions status
etc as solution 1.
Delay 1371 37 ps
Baseline Statistics
Baseline # W.Obs W.RMS Del N.R.D.
used/recov ps
GILCREEK-HARTRAO 14/ 14 68 1.01
GILCREEK-SESHAN25 127/ 131 46 .84
GILCREEK-TSUKUB32 366/ 383 30 .86
GILCREEK-WESTFORD 216/ 225 32 .79
HARTRAO -SESHAN25 52/ 55 46 .74
HARTRAO -TSUKUB32 53/ 53 44 .90
HARTRAO -WESTFORD 52/ 53 30 .84
SESHAN25-TSUKUB32 316/ 331 41 .87
SESHAN25-WESTFORD 39/ 40 55 .95
TSUKUB32-WESTFORD 136/ 141 43 .98
Corrected Reduced Chi-Square .8780
Solution with 4010/6010 KHz phases is slightly better than the solution with
10KHz phase cal phases, but improvement is marginal.
Full listing of the solution with 4010/6010
KHz phase cal tones.
Solution 3: with 4010/6010 KHz phase cal frequency
It has slighly different parameterization and weights than solution 2.
GILCREEK-HARTRAO 14/ 14 77 .88
GILCREEK-SESHAN25 128/ 132 45 .93
GILCREEK-TSUKUB32 366/ 383 30 .92
GILCREEK-WESTFORD 216/ 225 31 .93
HARTRAO -SESHAN25 52/ 55 44 .82
HARTRAO -TSUKUB32 53/ 53 46 .85
HARTRAO -WESTFORD 52/ 53 30 .78
SESHAN25-TSUKUB32 316/ 331 40 .93
SESHAN25-WESTFORD 39/ 40 54 .94
TSUKUB32-WESTFORD 136/ 141 43 .94
Solution 4: with 4010/6010 KHz phase cal frequency and with calibration
for spurious signals computed by Phase_Doctor:
It has exacatly the same parameterization, weights, suppressions status
etc as solution 3.
GILCREEK-HARTRAO 14/ 14 80 .91
GILCREEK-SESHAN25 128/ 132 44 .93
GILCREEK-TSUKUB32 366/ 383 30 .92
GILCREEK-WESTFORD 216/ 225 31 .92
HARTRAO -SESHAN25 52/ 55 42 .78
HARTRAO -TSUKUB32 53/ 53 45 .82
HARTRAO -WESTFORD 52/ 53 29 .75
SESHAN25-TSUKUB32 316/ 331 40 .93
SESHAN25-WESTFORD 39/ 40 54 .94
TSUKUB32-WESTFORD 136/ 141 43 .94
Spurious signals at 10KHz phase cal tones
Normalized phase cal amplitude versus phase cal phase. (Units for phases
at all plots are radians).
Spurious signals at 4010/6010 KHz phase cal tones
Normalized phase cal amplitude versus phase cal phase.
Differences in phase cal phases 4010/6010 KHz minus 10KHz tones tones
Conclusions
- Using 4010/6010 KHz phase cal marginally improved results of c1013
with respect to solution with 10KHz phases. Improvement is of about
the same magnitude as removal of spurious phase cal signals as
Phase_Doctor does it.
- Spurious signals in phase cal tones at 4010 KHz frequency is at about
the same level as spurious signals at 10 KHz frequency.
- Variations in the differences in phase cal phases 4010 KHz minus
10 KHz are at the level 0.03 - 0.30 rad ( 2 - 20 degrees ). They
depend on station and may show pattern and spikes. Points with
spikes have a tendency to be outliers in our solutions.
- Re-fringing old experiments is not justified.
- More investigation how and why phase of phase cal varies within
each individual channel is desirable. Extraction of all 8 phase
cal frequencies in rdv22 experiment promises to substantially
clarify situation.
Back to Leonid Petrov's discussion page.
This page was prepared by
Leonid Petrov
Last update: 12-DEC-2001 17:40:34