MeerKAT S band receiver commissioning (public part)

This area is for notes and results on the receiver commissioning. The namespace projects:mkat_sband is private to the MPIfR, the subspace projects:mkat__sband:pub has public read access.

SARAO web archive for commissioning proposal SSV-20200929-OW-01

Map of array

main

receptor status

S band install SN

Review of SARAO provided test data

S band main page

reference

All in MHz:

Please note that these are the total band ranges, the usable range is smaller. See description of the decimation filters for details.

A discussion of the digital decimation filters can be found here.

TO DO

List of data sets (automatically created with obs_info.py)

The data sets can be found in /fpra/polmag/01/MK_S_BAND_COM/MS/, some have copies (also of variants) in /media/storage_1/wucknitz/MEERKAT. Own conversion to MS via mvftoms.py have their capture block number as directory name, ones via the archive have e.g. 1603193556_sdp_l0.ms.

We determine delays and phases with fringe-fitting (click for details) per baseline. After calibrating the phases per baseline, we can measure the SNR (click for details) per baseline and channel and try to separate the baseline values to obtain station-based parameters. A similar approach is used for the bandpass (amplitude and phase, click for details).

Details of results for individual observations can be found here.

The performance of the entire array can be estimated best by taking the median (or some other percentile) over all antennas. Here are corresponding results from the 2022-11-19 observations (see below). First we take the median over all antennas and plot both polarisations individually:

Both polarisations

Second we average the polarisations but show the 10/50/90% percentile over all antennas. This means we see the typical range of antennas:

10/50/90% percentiles, polarisations averaged

In both files the first five panels show the individual subbands S0/1/2/3/4, the last one combines them all. The yellow vertical lines are the transition frequencies from one subband to the next. In reality the subbands are overlapping, but I always used the inner part where possible.

Here the same data are provided as text files with space-separated numbers. The first line in each file describes the fields (the same for all).

For a quick overview of the performance of all antennas, we provide curves for all antennas from relevant observations over all available subbands.

The first set of plots is meant to represent the latest measurements. These are currently the final SEFD measurements of 2022-11-19 at high elevations:

PDF with SEFD curves
PDF with bandpass curves
PDF with bandpass curves, amplitude corrected for digital decimation filter for a better comparison
PDF with SEFD and bandpass curves (dB)
PDF with SEFD and bandpass curves (dB), amplitude corrected for digital decimation filter for a better comparison

Note that the scales for some antennas are treated specially here for the moment, because the SEFD and bandpass phases are not in normal ranges.

The filter-corrected bandpass curves can be used to see how consistent the bandpass is between observations in different subband modes. This is possible because of a large overlap between the bands.

In order to study the differences between the two polarisations, we also provide plots of the logarithm of the radio pol0/pol1 for bandpass (signal, green), noise (SEFD, yellow) and signal-to-noise (red):

Ratio pol0/pol1 of signal, noise and signal/noise

Different signal levels for the two polarisation can move the green and yellow curves up or down together but would not affect the red one.

The following plots include all observations, which makes very large and difficult to view. These are mainly good for historical purposes:

PDF with SEFD curves
PDF with bandpass curves (linear)
PDF with bandpass curves (linear), amplitude corrected for digital decimation filter for a better comparison
PDF with SEFD and bandpass curves (dB)
PDF with SEFD and bandpass curves (dB), amplitude corrected for digital decimation filter for a better comparison

Here are the corresponding plots of the logarithm of the radio pol0/pol1 for bandpass (signal, green), noise (SEFD, yellow) and signal-to-noise (red):

Ratio pol0/pol1 of signal, noise and signal/noise (all obs)

A comparison of old and new decimation filters can be found here.

Receiver status (table at MPIfR)
new receiver status table (MPIfR)
Pieter Kotze's receiver status table

These are just rought numbers taken from the plots of the final measurements (only at the best frequencies). The replacement status is not necessarily up to date:

Three receivers/antennas (m015,m036,m062) are having problems at the moment. The others are all good, but have slightly different SEFD levels.

Analysis of several types of system noise variations