Pulsars

Introduction

The VLA is capable of performing imaging observations of pulsars by "phase-binning" visibilities (also sometimes known as gating).  In this observing mode, rather than continuously integrating visibility data for several seconds as in a standard observation, the data are accumulated into a number of "bins" corresponding to different rotational phase ranges of the pulsar.  The predicted phase versus time is computed from a user-supplied pulsar timing ephemeris; see below for more details.  This allows each pulse phase bin to be imaged separately, for example to identify which continuum source in a given field is the pulsar, or to image the surrounding area after subtracting off pulsed emission, or simply to increase S/N on the pulsar relative to a standard image, etc.

Note that this mode is not suitable for observations of pulsars whose period is unknown a priori, or for observations of single pulses from sources such as FRBs or RRATs.

Correlator Setup

There are several constraints on the bin width, total allowed bandwidth, and data rate that are specific to pulsar binning observations.  These are described in detail in the OSS pulsar section and in the VLA proposal guide.  Once you have decided on your basic setup (band, frequency resolution, etc), set up the subbands following the general instructions given in the RCT manual.  Pulsar binning can then be enabled by going to the "Pulsar" tab in the RCT, selecting "Phase Binned Imaging" mode, and entering the number of bins, as shown below:

The final option specifies whether the pulse period model should be computed from a timing ephemeris, also known as a "par file" or alternately if the binning should be done at a constant topocentric pulse period.  The RCT will verify that the chosen configuration is possible, and will report the total data rate, as shown.

Calibration

In general there is no special calibration that needs to be done for binning observations.  Calibration procedures for complex gain, flux density scale, and if necessary polarization should follow the general guidelines laid out in the Calibration section of the manual.  The only notable differences for pulsar observing are:

1. Calibrators should be observed using a standard (non-binned) correlator setup.  It is recommended that this setup use the same frequency configuration (tuning frequencies, subbands, numbers of channels) as is used for your pulsar target.  The easiest thing to do is first set up the pulsar configuration, then make a copy of it in the RCT and disable binning.
2. As for all VLA observing, binning projects should include an initial setup scan for each band used, to set attenuator levels.  However, there is no need to include additional setup scans to trigger requantizer setting.

Pulsar Ephemerides

When observing pulsars the apparent pulse period is slowly changing with time due to a number of effects including motion of the Earth, motion of the pulsar (binary orbits), intrinsic spin evolution, etc.  The period versus time can be predicted using an ephemeris or timing model.  For binning observations, you should supply an ephemeris using the standard TEMPO "par file" format.  A polynomial approximation to the pulse phase versus time will be generated on the fly using TEMPO at the time of your observation and used for binning the visibilities.  Note that TEMPO2-format par files are not currently accepted.  If you have a par file in TEMPO2 format (for example if it includes "UNITS TCB"), you will need to convert it to TEMPO (TDB) format.

Pulsar ephemerides should be uploaded into your source catalog in the SCT via the "Pulsar Ephemeris" tab:

Since VLA binning observations always generate bins covering the full pulse period there is no need to have absolute phase information about the source you are observing.  This also means it is not strictly necessary that the ephemeris be accurate over the full length of your observation, although the pulse "drifting" into different bins versus time may make data analysis harder.  The fundamental requirement to avoid smearing out pulses is that ΔP x δt < 1/N_bin, where ΔP is the error in the predicted pulse period and δt is the correlator integration time.

Many pulsar ephemerides can be found in the ATNF Pulsar Catalog, but they are not necessarily up to date for all sources.  It is always worthwhile to verify a timing model with recent data if there is any uncertainty about it.