Westerbork TOG report

Hardware

New Tied Array (TADUmin)

The new tied array was used for VLBI sessions in November 2002, February and May/June 2003. Shortly before the November session a bug in the specification was removed, and between Feb and May the output stage equalizers and filters were modified. This was introduced with a minor change in mixing scheme.
The new tied array is a hybrid of digital adding system with 8 pairs of analog outputs. These go through a series of mixers (with 4 LOs) to produce an IF band of 103-257MHz, suitable for VLBI input. The IF does not reproduce the input sky frequency, and so cannot be represented by a single effective LO.

Mark5 tests and formatter ID

During this period we have used Mark5 recorders borrowed from JIVE to perform tests and do observations. This is awkward as we our local (Westerbork) network force us to change IP address and socket numbers. During the May/June VLBI Mk5 tests we stumbled with the hardware ID of our formatter. We had no pins to re-jumper the number and so have (after the session) hardwired a new number inside the formatter.

Software

Linux kernel

In December we upgraded our linux kernel to 2.4.18 (Debian Woody 3.0). This was done so that we could support newer hardware for our backup FS computer. The upgrade used the tips provided by Pablo de Vicente.
We had some problems with automated scripts to send clock offset information to Bologna

FS

We have continued to upgrade FS versions between VLBI sessions and at present use fs-9.6.2

System

17MHz ripple

We are at present trying to cure a 17MHz standing wave 'ripple' in all observing bands.

WSRT uv data

The new tied array is calibrated using a line backend, so by default all observations will be done with 64 channels per 16MHz band. This data will be available after observing, but PIs who want it should request it in advance.

Future developments

Mark 5

Westerbork has ordered 2 Mark5 VLBI recorders from Conduant, but not yet ordered the disks- we want advice on what the best buy at the time is (maybe 300GB disks?)

eVLBI

The single-mode optical fibers between Westerbork and Dwingeloo already exist, and will soon be lit. The plan is to have one pair for ordinary routed connections and a separate pair for dedicated eVLBI use. The dedicated eVLBI system will not be routed, but come directly to JIVE. At present our network technician is planning to use Allied Telesyn PCI cards to do this, as the distance is only about 27km.

Methanol line receiver

The 5cm receiver is partially assembled. It will replace a normal MFFE receiver package for single-dish methanol VLBI. All frequency changes will be via operator control

VHF (115 to 175MHz)

We have plans for an even lower frequency band, to allow correlations between Westerbork and LOFAR. The funding for these is expected soon; the main problem is the filtering and high dynamic range needed, but the physical size is also problematic.

Fully digital tied array (TADU)

There is now a plan to upgrade to a fully digital tied array. This should support both pulsar and VLBI operations, and the digital output will avoid the IF rack completely and by connected to the formatter sampler output. The connections of sky frequency and polarizations will be handled in a crossbar switch. Some of the hardware will be compatible with the Mk4 correlator; serial-link systems are being re-implemented in newer hardware with a copper or optical fiber option. It is planned to make it VSI-H compatible. We hope that the TADU will be available by 2005.