In all receivers the second polarisation is systematically worse than the first one. In m005 this effect is even stronger than in the others.

In the spectral plots the grey areas have been excluded from the calibration fits due to RFI or aliasing effects near the band edges. The dashed curves in the baseline plots are re-derived from the antenna-based solutions. They are consistent with the original parameters, which proves the absence of significant closure errors.

The spectral information is generally binned for the plots by a factor of 4 or 32 for final 1k channels. For the calibration we generally used the 10-sec solutions with group delay and without explicit curvature. Antenna m008 was excluded, because it has problems with the synchronisation and did not produce correlated signals.

Lines are not labeled in all plots. Generally red is for the first polarisation (0 or H), green for the second (1 or V).

used baseline lengths (xy) > 250 m

calibration solutions:
baseline-based
antenna-based

SEFD and (amplitude and phase) bandpass:
baseline-based
antenna-based

linear gradient fits for delay solutions compared to observed solutions:
antenna-based dispersive delay
antenna-based non-dispersive delay
baseline-based dispersive delay
baseline-based non-dispersive delay
As described in Fringe-fitting, the dispersive delay is pretty consistent with a time-varying gradient across the array.

• 23-29 deg elevation
• antennas included: m000 m002 m003 m004 m005 m006 m007 (m008) m030 m045 m053 m057 m060
• m008 not working (excluded from fits)
• m004, m060 very bad SEFD, differs between pols, m005 also not very good
• Strong ripples and fluctuations in SEFD curves and bandpasses, different from antenna to antenna. 'Resonance' near 3065 MHz particularly concpicuous.
• Best SEFD near 420 Jy.
• 4k mode works, S3 works

used baseline lengths (xy) > 250 m

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

linear gradient fits for delay solutions compared to observed solutions:
antenna-based dispersive delay
antenna-based non-dispersive delay
baseline-based dispersive delay
baseline-based non-dispersive delay
Again the dispersive delay is pretty consistent with a time-varying gradient across the array.

• 36-42 deg elevation
• antennas included: m000 m002 m003 m004 m005 m006 m007 (m008) m030 m045 m053 m057 m060
• m004, m060 very bad SEFD, differs between pols, m005 also not very good
• m008 not working (excluded from fits)
• Strong ripples and fluctuations in SEFD curves and bandpasses, different from antenna to antenna. 'Resonance' near 3065 MHz particularly concpicuous.
• Best SEFD near 400 Jy.
• 32k mode works, S3 works

used baseline lengths (xy) > 300 m

Because m056 has problems in the second half of the observations with the second polarisation, I produced two versions of the results. The first is without m056 and uses the entire observation:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

linear gradient fits for delay solutions compared to observed solutions:
antenna-based dispersive delay
antenna-based non-dispersive delay
baseline-based dispersive delay
baseline-based non-dispersive delay
Here the picture is less clear, the ionospheric effects are not just simple large-scale gradients.

The second uses all antennas (with the exception of m008) and only the first 30 min:

Baseline-based calibration solutions
Antenna-based calibration solutions

Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

In an earlier version I found strange peaks in the solutions, some of which were not even consistent between the polarisations. This was an artefact of residual RFI and a problem with my fitting code. Now the solutions look stable and normal.

• 54-57 deg elevation
• antennas included: m000 m002 m003 m004 m005 m006 m007 (m008) m030 m045 m053 m056 m060
• m004, m060, m056 very bad SEFD, differs between pols, m005 also not very good
• m056 second polarisation has problems in second half (has to be discussed elsewhere), this is confirmed by Pieter Kotze ("M056 V channel appear to be 'noisy' at times.")
• m008 not working (excluded from fits)
• Strong ripples and fluctuations in SEFD curves and bandpasses, particularly in upper half of the band, different from antenna to antenna.
• Best SEFD slightly below 400 Jy.
• 4k mode works, S4 works

used baseline lengths (xy) > 300 m

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

linear gradient fits for delay solutions compared to observed solutions:
antenna-based dispersive delay
antenna-based non-dispersive delay
baseline-based dispersive delay
baseline-based non-dispersive delay
The ionospheric effects are not just simple large-scale gradients.

• 21-27 deg elevation
• antennas included: m002 m003 m004 m005 m006 m007 m021 m030 m043 m049 m053 m056 m060
• m056 is again 'noisy' and excluded from analysis
• m043 bad, particularly in second pol
• Best SEFD slightly below 420 Jy.

used baseline lengths (xy) > 300 m

Short test after replacing dipoles in m004.

Because m001 has very low S/N, which affects the station-based solutions also for the other stations, we have two sets of results.

Without m001:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

With m001, really only good for that station:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

In the collected plots, we use the former for all but m001 and the latter for m001.

• 39-41 deg elevation
• antennas included: m001 m002 m003 m004 m005 m006 m007 m013 m018 m021 m030 m049 m053 m057
• no signal in m013 and m018
• m001 extremely bad SEFD (note different scale for this one in the plots), probably not ready yet
• Best SEFD near 400 Jy.

used baseline lengths (xy) > 300 m

Short test after replacing LNAs in m004 (in addition to dipoles), dipoles in m043, m056, m060, one LNA in m043, trying to get m008 to sync.

Because m001 and m048 have very low S/N, which affects the station-based solutions also for the other stations, we have two sets of results.

Without m001 and m048:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

With m001 and m048, really only good for those stations:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

In the collected plots, we use the former for all but m001 and m048 and the latter for m001 and m048.

• 45-46 deg elevation
• m043, m056, m060 have improved a lot (second pol)!
• antennas included: m000 m001 m002 m003 m005 m006 m007 m030 m043 m048 m049 m053 m056 m057 m060
• m001 and m048 extremely bad SEFD (note different scale for this one in the plots), probably not ready yet
• Best SEFD near 400 Jy.

used baseline lengths (xy) > 300 m

Short test after LNAs in m004 and one in m043, dipoles in m043, m056, m060, trying to get m008 to sync.

Because m001 and m048 have very low S/N, which affects the station-based solutions also for the other stations, we have two sets of results.

Without m001 and m048:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

With m001 and m048, really only good for those stations:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

In the collected plots, we use the former for all but m001 and m048 and the latter for m001 and m048.

• 31-38 deg elevation
• antennas included: m000 m001 m002 m003 m005 m006 m007 m018 m022 m030 m038 m043 m048 m049 m053 m056 m057 m060
• no signal in m018, m022, m038
• Best SEFD near 400 Jy.

used baseline lengths (xy) > 300 m

Short test after LNAs in m004 and one in m043, dipoles in m043, m056, m060, trying to get m008 to sync.

Because m001 and m048 have very low S/N, which affects the station-based solutions also for the other stations, we have two sets of results.

Without m001 and m048:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

With m001 and m048, really only good for those stations:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

In the collected plots, we use the former for all but m001 and m048 and the latter for m001 and m048.

• 57 deg elevation
• new antenna m001: good SEFD
• m004 now much better (both pols)
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m008 m011 m021 m030 m043 m048 m049 m053 m056 m057 m060
• still no signal in m008
• Best SEFD near 400 Jy.

• m008 still no signal, probably not in sync
• m011 (new receiver) looks good
• m001 still bad, m048 (new receiver) same
• both bad pols improved significantly in m004
• bad second pols improved significantly in m043, m056, m060

First attempts of subbands S0-2, plus S3 for comparison. Also additional antennas.

used baseline lengths (xy) > 300 m

Because m001 m013 m018 m022 m048 m050 m054 m061 have very low S/N, which affects the station-based solutions also for the other stations, we have two sets of results, one with these antennas included (used only for them in the collected plots) and one with these excluded.

Without m001 m013 m018 m022 m048 m050 m054 m061:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

Without m001 m013 m018 m022 m048 m050 m054 m061:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

Without m001 m013 m018 m022 m048 m050 m054 m061:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

Without m001 m013 m018 m022 m048 m050 m054 m061:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

Without m001 m013 m018 m022 m048 m050 m054 m061:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

Without m001 m013 m018 m022 m048 m050 m054 m061:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

Without m001 m013 m018 m022 m048 m050 m054 m061:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

Without m001 m013 m018 m022 m048 m050 m054 m061:

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

• 30-47 deg elevation
• antennas included: m000 m001 m003 m004 m005 m011 m013 m018 m021 m022 m027 m030 m038 m043 m045 m048 m049 m050 m053 m054 m056
• (still) low SNR in m001 m013 m018 m022 m048 m050 m054 m061
• new good antennas: m027 m037, new bad antennas: m013 m018 m022 m050 m054 m061
• now tried all subbands, all seem to work
• more RFI in lower subbands

used baseline lengths (xy) > 300 m

All antennas included deliver good or at least usable signal.

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

• 55-57 deg elevation
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m008 m011 m013 m018 m022 m027 m030 m038 m043 m045 m049 m050 m053 m054 m056 m057 m059 m060 m061 m063, all deliver usable signal
• in S4 now worse than before: m004 (400 → 550 Jy), m005 (380 → 500 Jy), m006 (400 → 550 Jy), m027 (400 → 500 Jy), m030 (400 → 550 Jy) m045 (390 → 520 Jy), m053 (450 → 500 Jy), m057 (400 → 500 Jy), m060 (390 → 530 Jy)
• now good/okay: m001 m013 m022 m061
• now good/okay (no data before): m008, m018 (first pol bad), m063, m059, m063
• m038, m050, m054 improved

used baseline lengths (xy) > 300 m

Most antennas included deliver good or at least usable signal, no signal from m025, very bad from m014 (both not included in the plots).

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

• 51-52 deg elevation
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m008 m009 m010 m011 m012 m013 m014 m019 m025 m026 m027 m030 m032 m033 m034 m035 m038 m043 m045 m048 m049 m050 m053 m054 m056 m057 m059 m060 m063, most deliver usable signal
• no signal at all from m025
• extremely bad signal (basically no sensitivity) from m014
• new antennas delivering good signal for first time: m009 m010 m012 m019 m026 m032 m033 m034 m035
• m048 now good for the first time

used baseline lengths (xy) > 300 m

All antennas included deliver good or at least usable signal.

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

• 49-54 deg elevation
• antennas included: m001 m002 m003 m004 m005 m006 m007 m008 m009 m010 m011 m012 m013 m014 m017 m018 m019 m020 m021 m022 m023 m027 m030 m032 m034 m035 m038 m043 m045 m048 m049 m050 m053 m054 m056 m057 m059 m060 m061 m063, almost all deliver good signal
• no usable signal from m018 (was better earlier)
• m014 now good
• new good data: m017 m020 m023 (second pol worse in m020 m023)
• many receivers have improved, including ones that first were good and then got worse
• problem with noise diode was apparently responsible for bad SEFD in many receivers, this problem is solved

used baseline lengths (xy) > 300 m

Almost all antennas included deliver good or at least usable signal. Very low for m020 and m038, those not included in analysis.

Cannot download the file, the conversion to MS got stuck.

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

linear gradient fits for delay solutions compared to observed solutions:
antenna-based dispersive delay
antenna-based non-dispersive delay
baseline-based dispersive delay
baseline-based non-dispersive delay
See Fringe-fitting for a description of this analysis. The dispersive delay is consistent with a time-varying gradient across the array, the non-dispersive delay not. This is all as expected from the ionosphere and troposphere.

• 28-34 deg elevation
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m008 m009 m011 m012 m013 m014 m017 m019 m020 m021 m022 m023 m030 m031 m032 m034 m035 m038 m043 m045 m048 m049 m050 m054 m056 m057 m059 m060 m061 m063, almost all deliver good signal
• no good signal from m020, m038
• first good signal from m031: SEFD ~ 430 Jy
• m008: second pol worse in lower part of the band
• phases very stable over one hour, phases and delays fully consistent with troposhere and ionosphere, potential delay trends certainly < 50 picosec (1.5cm), probably much smaller

used baseline lengths (xy) > 300 m

Most antennas included deliver good or at least usable signal, no signal from m025, very bad from m014 (both not included in the plots).

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

• 24-25 deg elevation
• antennas included: m000 m001 m002 m003 m004 m005 m007 m008 m009 m011 m012 m013 m014 m017 m019 m021 m022 m023 m024 m026 m027 m030 m031 m032 m033 m034 m035 m037 m038 m043 m045 m048 m053 m054 m056 m057 m059 m060 m061, all deliver usable signal
• new antennas delivering good signal for first time: m024 m037
• elevation is lower than e.g. in 1615522969, as result of that the SEFD is higher now, but strangely only by 50 Jy for some antennas but by 200 Jy for others

used baseline lengths (xy) > 300 m

This is one scan (starting ca. 13:59 UTC) of PKS 1934-63 extracted from the 32k test dataset. It was analysed to get an up-to-date view of the receiver performance.

Most antennas included deliver good or at least usable signal, no signal from m058 (not included in plots), very bad from m031.

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass

Note that the data weights are now scaled differently than in raw MS and were therefore not used as noise estimate. The difference between frequency channels that was used instead is much more noise (therefore no smooth SEFD curves).

• 37 deg elevation
• antennas included: m000 m001 m003 m004 m005 m006 m007 m009 m010 m012 m014 m015 m016 m017 m018 m019 m020 m021 m022 m023 m024 m025 m026 m030 m031 m032 m033 m034 m035 m036 m037 m038 m039 m040 m041 m042 m043 m044 m045 m047 m048 m049 m050 m051 m053 m054 m056 m057 m058 m059 m061 m062 m063
• 'm005': second pol worse than first
• 'm006': relatively high SEFD
• 'm012': second pol worse than first
• 'm015': second pol much worse than first
• 'm019': relatively high SEFD, strong dependence on frequency
• 'm020': second pol worse than first
• 'm022': second pol worse than first
• 'm023': second pol worse than first
• 'm025': second pol worse than first
• 'm031': very low signal
• 'm034': second pol worse than first
• 'm041': second pol worse than first
• 'm042': second pol worse than first, strong bumps in SEFD
• 'm047': second pol worse than first, strong bumps in SEFD
• 'm051': second pol worse than first
• 'm054': very low signal, strange behaviour with frequency
• 'm059': second pol worse than first

used baseline lengths (xy) > 300 m

All antennas included deliver good or at least some signal. Very low for m031 and m054.

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass, auto-limits in SEFD curves so that ''m031'' and ''m054'' are no longer outside of the plotting rage

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass, auto-limits in SEFD curves so that ''m031'' and ''m054'' are no longer outside of the plotting rage

• 34-47 deg elevation
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m020 m021 m022 m023 m024 m025 m026 m027 m028 m029 m030 m031 m033 m034 m035 m037 m041 m042 m043 m044 m045 m047 m049 m050 m051 m053 m054 m055 m056 m057 m058 m059 m061 m062 m063 , almost all deliver good signal
• no good signal from m031, m054

used baseline lengths (xy) > 300 m

All antennas included deliver good or at least some signal. Very low for m031.

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass, auto-limits in SEFD curves so that ''m031'' is no longer outside of the plotting rage

• 48-49 deg elevation
• antennas included: m001 m002 m003 m004 m005 m006 m007 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m023 m024 m025 m026 m027 m028 m029 m030 m031 m032 m033 m034 m035 m036 m037 m038 m040 m043 m044 m045 m047 m048 m050 m051 m053 m054 m055 m056 m057 m058 m059 m060 m061 m062 m063, almost all deliver good signal
• m006 better
• m031 still very bad
• m036 new, good!
• m041 new, first pol good, second worse
• m054 now good!

used baseline lengths (xy) > 300 m

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass, auto-limits in SEFD curves so that ''m031'' is no longer outside of the plotting rage

• 21-22 deg elevation
• no signal from m007, m031 still very bad, still bump in m015
• antennas included: m000 m001 m003 m005 m007 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m023 m024 m025 m026 m027 m029 m030 m031 m032 m033 m034 m035 m037 m038 m041 m042 m043 m044 m047 m048 m049 m050 m051 m053 m054 m055 m057 m059 m060 m061 m062 m063, almost all deliver good signal

used baseline lengths (xy) > 300 m

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass, auto-limits in SEFD curves so that ''m031'' is no longer outside of the plotting rage

• 26 deg elevation
• no signal from m007, m031 still very bad, still bump in m015
• antennas included: m000 m001 m003 m005 m007 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m023 m024 m025 m026 m027 m029 m030 m031 m032 m033 m034 m035 m037 m038 m041 m042 m043 m044 m047 m048 m049 m050 m051 m053 m054 m055 m057 m059 m060 m061 m062 m063, almost all deliver good signal

used baseline lengths (xy) > 300 m

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass, auto-limits in SEFD curves so that ''m031'' is no longer outside of the plotting rage

• 33 deg elevation
• no good signal from m031
• antennas included: m000 m001 m002 m003 m004 m005 m006 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m025 m026 m027 m029 m030 m031 m032 m034 m035 m036 m037 m038 m039 m040 m041 m042 m043 m044 m046 m048 m049 m051 m053 m054 m055 m056 m057 m058 m060 m061 m063

used baseline lengths (xy) > 300 m

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass, auto-limits in SEFD curves so that ''m031'' is no longer outside of the plotting rage

• 33 deg elevation
• no good signal from m031 and m007
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m025 m026 m027 m029 m030 m031 m032 m034 m035 m036 m037 m038 m039 m040 m041 m042 m043 m044 m046 m048 m049 m051 m053 m054 m055 m056 m057 m058 m060 m061 m063

This is a phase stability test. The SEFD and gain curve results are not included in the general overviews.

Data were averaged to 10-sec in the conversion to MS. Solutions are for individual 10-sec blocks and therefore without rates.

used baseline lengths (xy) > 300 m

calibration solutions:
baseline-based
antenna-based

SEFD and (amplitude and phase) bandpass:
baseline-based
antenna-based
Please note that the SEFD results should not be taken too seriously, because the way I compute the SNR is affected by the averaging in time.

linear gradient fits for delay solutions compared to observed solutions:
antenna-based dispersive delay
antenna-based non-dispersive delay
baseline-based dispersive delay
baseline-based non-dispersive delay
See Fringe-fitting for more explanations on these fits.

Here are plots that simulate the phasing efficiency. It is assumed that the delays are calibrated on the target at the beginning (top) or centre (middle) or end (bottom) of the observation. Different numbers of antennas (starting at the core) are included (several curves). These efficiencies are for the actually used observing frequency. Because I have dispersive and non-dispersive effects separated, I can extrapolate these results to other frequencies, if needed.

This is for narrow-band observations in the centre of the subband. I could include bandwidth effects, but they are not significant at these levels.

Here is the same, but scaled to 600 MHz frequency (lower end of UHF). The dispersive and non-dispersive effects are of course scaled separately.

Results of a more general statistical analysis can be found in 1636719400_phaseff_obsfreq.pdf. Each page is for a selection of antennas. The upper panels show the phasing efficiency for each time bin calibrated with each other time bin. The lower panel show this as function of time lag between calibration and target time (percentiles for certain fractions of the time). Not that the number of samples is small on the right end of this plot, so it is less statistically meaningful there.

The same is available scaled to 600 MHz in 1636719400_phaseff_600mhz.pdf.

Note that the phases are more stable in UHF than S band. This is because in S band the ionosphere is only a very minor contribution. The troposphere scales down at lower frequency. The ionosphere becomes more important but still is not too problematic. The behaviour is slightly different at long baselines, where the ionosphere is comparatively stronger, because its gradients extend over longer distances.

• 49-57 deg elevation
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m023 m025 m026 m027 m029 m030 m031 m032 m033 m034 m035 m036 m037 m038 m039 m040 m041 m042 m043 m044 m045 m046 m047 m048 m049 m050 m051 m053 m054 m055 m056 m057 m058 m060 m061 m062 m063
• not included in the analysis because of relatively low SNR: m002 m007 m026 m031 m033 m060, not included because of intermittent ampltitude dip (tracking problem?): m009
• The linear gradient fits show that the dispersive delays are consistent with an ionosphere that almost entirely consists of a time-varying gradient across the array. For the non-dispersive delay the atmospheric structures seem to be much smaller, as expected. There are no strong long-term trends and the behaviour does not look unusual for atmospheric effects. We could produce structure functions for different baselines, but those only help if we have a similar analysis for the expected atmosphere (or for L band).

used baseline lengths (xy) > 300 m

calibration solutions:
Baseline-based calibration solutions
Antenna-based calibration solutions

SEFD and (amplitude and phase) bandpass:
Baseline-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass
Antenna-based SEFD and amplitude+phase bandpass, auto-limits in SEFD curves so that ''m031'' is no longer outside of the plotting rage

• 57 deg elevation
• m002: better, difference between pols reduced
• m005: not better, maybe even worse
• m015: not better, maybe even worse, still with bump in second pol
• m026: maybe slightly worse, difference between pols reduced
• m031 still very bad (around or above 20kJy)
• m034: worse
• m055: worse
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m023 m024 m025 m026 m027 m029 m030 m031 m032 m033 m034 m035 m036 m037 m038 m040 m041 m042 m044 m045 m046 m047 m049 m050 m051 m053 m055 m057 m058 m059 m060

This is a phase stability test. The SEFD and gain curve results are not included in the general overviews.

Data were averaged to 10-sec in the conversion to MS, XY and YX are discarded. Solutions are for individual 10-sec blocks and therefore without rates.

used baseline lengths (xy) > 300 m

This time the baseline-based plots are not included, because they take too much space on the wiki server.

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based
Please note that the SEFD results should not be taken too seriously, because the way I compute the SNR is affected by the averaging in time.

linear gradient fits for delay solutions compared to observed solutions:
antenna-based dispersive delay
antenna-based non-dispersive delay

See Fringe-fitting for more explanations on these fits.

• 52-55 deg elevation
• antennas included: m001 m002 m003 m004 m005 m006 m007 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m023 m024 m025 m026 m027 m028 m029 m030 m031 m032 m033 m034 m035 m037 m038 m040 m041 m042 m044 m045 m046 m047 m048 m049 m050 m051 m053 m055 m056 m057 m058 m059 m060 m061 m062 m063
• m031 is now looking good
• not included in the analysis because of lower SNR and difference between X and Y sensitivity: m015, m034
• The linear gradient fits show that the dispersive delays are consistent with an ionosphere that almost entirely consists of a time-varying gradient across the array. For the non-dispersive delay the atmospheric structures seem to be much smaller, as expected. There are no strong long-term trends and the behaviour does not look unusual for atmospheric effects.
• There are stronger short-term fluctuations than in 1636719400. Was the weather different?
• There are periods with increased wide-band noise. Not sure if this is due to an active Sun. Interestingly the amplitude solutions are lower by a few per cent at those time. Maybe the levels are too high?

Here are plots that simulate the phasing efficiency. It is assumed that the delays are calibrated on the target at the beginning (top) or centre (middle) or end (bottom) of the observation. Different numbers of antennas (starting at the core) are included (several curves). These efficiencies are for the actually used observing frequency. Because I have dispersive and non-dispersive effects separated, I can extrapolate these results to other frequencies, if needed.

This is for narrow-band observations in the centre of the subband. I could include bandwidth effects, but they are not significant at these levels.

Here the same for 600 MHz:

And here the statistics for the observed frequency and for 600 MHz: 1638617387_phaseff_obsfreq.pdf 1638617387_phaseff_600mhz.pdf

The new filter coefficients were installed on 1st June 2022. On 8th tests were run in each of the subbands. Dump time was 8 sec, each observation had 10 min. Only baseline lengths (xy) > 300 m were used. The antenna set was the same for all five observations.

The baseline-based plots are not included, because they take too much space on the wiki server. They can be provided on request.

elev. 24 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

elev. 27 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

elev. 31 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

elev. 34 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

elev. 37 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

• 24-37 deg elevation
• antennas included: m000 m001 m002 m003 m004 m005 m006 m007 m009 m010 m011 m012 m013 m014 m015 m016 m018 m019 m020 m021 m022 m023 m024 m026 m027 m028 m029 m030 m031 m032 m033 m034 m035 m036 m037 m038 m039 m041 m042 m044 m045 m046 m047 m049 m050 m051 m052 m053 m054 m057 m058 m059 m060 m061
• second pol bad in m015, first pol bad in m020, both pols not great in m034
• the filters are behaving as expected
• Combined results and a comparison with the old filters can be found here.

Short 5-min observations of PKS1934-63 were done at high elevation on 19th November 2022 in each of the five subbands. Dump time was 2 sec, solution interval for the calibration was 10 sec. Only baseline lengths (xy) > 300 m were used for the analysis.

The baseline-based plots are not included, because they take too much space on the wiki server. They can be provided on request.

elev. 56 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

elev. 54 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

elev. 57 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

elev. 57 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

elev. 57 deg

calibration solutions:
antenna-based

SEFD and (amplitude and phase) bandpass:
antenna-based

• 54-57 deg elevation
• antennas included: m000 m001 m002 m005 m006 m007 m008 m009 m010 m011 m012 m013 m014 m015 m016 m017 m018 m019 m020 m021 m022 m023 m024 m025 m026 m027 m028 m029 m030 m031 m032 m033 m034 m035 m036 m037 m038 m039 m040 m041 m042 m044 m045 m046 m047 m048 m049 m050 m051 m052 m053 m054 m055 m057 m058 m059 m060 m061 m062 m063 (m036 m037 missing in S1)
• results generally very consistent (but somewhat variable) from antenna to antenna
• m015: second pol still bad (huge bumps e.g. around 2900 MHz)
• m036: second pol very bad
• m062: first pol has large bumps, e.g. just above 2200 MHz
• Combined results can be found here.