Two different antenna are used for observing at ASKAP - AK07 has an L-band feed permanently installed while AK32 has an X-band feed installed. Both are connected to the control room by RF over fibre technology and fed into an RF switch to allow recording from one of the telescopes.
There are 3 main processes run for ASKAP observing. askap_drv is the main telescope control. It itself consists of a terminal based control process and a GUI. askap_drv must be running. VLBI scheduling is run from a process called “vlbi_drv”, while the actual recording is via a command line process called px14_record. Generally VLBI observers are not expected to start any of these processes, though they do crash/freeze from time to time.
See the test screencast covering these notes at:
Note that the antenna control and recording (px14_record) are totally separate. You can restart one without affecting the other.
Make sure previous instances of askap_drv and vlbi_drv are not running. ^C any stalled processes if necessary.
The driver task “askap_drv” needs to run first for communicating with the antenna. This can be done by:
> cd askap/run > askap #n
where #n is the CETCI antenna number. 26 for Ak32 (X-band) or 19 for Ak07 (L-band). e.g. “
askap 26” to run with AK32.
This starts askap_drv and the antenna gui in one step.
If the antenna is stowed (“
high_stow” in and red), then type “
unstow” into the gui command input. Other commands you may need are:
|stow||Stow the antenna|
|unstow||Unstow the antenna|
|drive_off||Manually turn drives off|
|drive_on||Manually turn drives on|
|update||Synch with the ntp server|
To run a VLBI schedule, use
> cd askap/run > vlbi_drv ../sched/vt999a-ak.psn
vt999a with the name of the experiment). This assumes schedules have been previously copied to ~vlbi/askap/sched.
Data is recorded using the program px14_record. It is a standalone terminal based program which reads data off the installed PCI sampler card and writes formatted data to local disk (or eVLBI network streaming).
There are currently (24/3/2015) 3 data disks -
/mnt/raid_0, /mnt/raid_1, /mnt/raid_2. You need to manually “
cd” to a data disk with enough free space and create a directory for the experiment. px14_record can be run as
> px14_record -o <EXPER> -t <TT>h -v <VV>
|<EXPER>||File name prefix - usually the experiment name|
|<TT>||Recording time. 12h is 12 hours, 10s is 10 seconds|
|<VV>||Internal gain -usually about 10 for L-band, 20 for X-band|
|-invert||Invert the spectra, needed if using an effective lower sideband mixing|
An example usage is:
> df -h > cd /mnt/raid_1 > mkdir vt999a > cd vt999a > px14_record -o vt999a -t 10h -v 12
Typical output is:
1.0 nbuf=4646 Rate=514 Stddev=1123.7,1569.4 Elapsed 00:00:19 vt999a_083_021047.lba -15.87
This shows time since last update (1.0 sec), number of buffers recorded, approx recording data rate, StdDev of the two polarisations (from 12 bit data), Elapsed time since starting, current file name and clock offset in millisec.
-v” option to increase or decrease gain/attenuation on the sampler card
px14_recordcan be terminated either typing “escape” or ^C. Escape is “safer”.
Regularly (10-15min) check the following. The drive system on ASKAP is known to fail moderately regularly
drive_off”. Wait a few seconds. type “
drive_on”. If the errors do not go repeat the drive off/on sequence a few more times.