View page
Talk page

Observations: GMVA ALMA mode

Modified 15:14, 29 Mar 2017 by rottmann | Page History

    Version as of 09:01, 27 Nov 2024

    to this version.

    Return to Version archive.

    View current version

    Prerequisits

    • Go through the system-setup and make sure the system is operational.
    • Assumption: Fibre cable is connecting the FiLA10G to eth2 of pv-mark6-1
    • A single module is loaded and connected into slot 1 of pv-mark6-1
    • Make sure you have loaded the  module for recording the GMVA ALMA mode

    FiLA10G Setup

    • On the DBBC2 desktop double-click the icon labeled "fila10g_ALMAMode.bat". This will setup the FiLA10G for recording on the Mark6, will do the time-synchronisation and will do the required channel re-ordering of the VDIF-data. The icon is a shortcut  to the following file: C:\DBBC\bin\fila10g_AlmaMode.bat.

     

    • Confirm settings. In the terminal window that has opened the last page lists the output of the sysstat command. Inspect the output; in particular check that VDIF timesync is showing "yes" and "Output" is showing "started" (see image below).

    sysstat_mark6.png

    Mark6 module initialization

    • Skip this section if there is already recorded data on the module! The following steps will delete all data on the module!
    • log-into pv-mark6-1 as user oper
    • execute: da-client
    • in da-client: 
      mod_init=1:8:{VSN}:sg:new           replace    {VSN} with VSN as indicated on the module label
group=new:1

    Mark6 setup

    • log-into pv-mark6-1:     ssh pv-mark6-1 -l oper -X
    • execute:     m6service_restart
    • execute:     da-client
    • in da-client:
    mstat?1

    make sure the group is NOT in open state. If it is open you must do:

    group=close:1

    input_stream=add:DBBC2:vdif:5032:50:42:eth2:::1
input_stream=commit
group=open:1
mstat?1

    The last statement should report the group to be in open:ready state, e.g.:

    !mstat?0:0:1:1:MPIH%016/48008/4/8:8:8:45160:48008:open:ready:sg;

    Loading/starting the schedule

    • log-into pv-mark6-1 as user oper
    • Download/copy the vex file into the directory: /home/oper/vex
    • Translate the vex-file
    cd /home/oper/vex
vex2xml.py -f {vexfile} -s PV
    • Check the contents of the produced xml-file. It should not be empty and contain several scan entries, similar to this:
    <experiment name="a90xx" station="PV" start="2015090003000"  end="2015090032200">
        <scan experiment="a90xx" source="3C279" station_code="PV" start_time="2015090003000" duration="420" scan_name="090-0030"/>
        <scan experiment="a90xx" source="3C279" station_code="PV" start_time="2015090004500" duration="420"
        <scan experiment="a90xx" source="3C279" station_code="PV" start_time="2015090031500" duration="420" scan_name="090-0315"/>
</experiment>

    NOTE: If the file is empty check that the station code given in the vex2xml.py command matches the station code for Pico Veleta of the vex-file.

    • Start the schedule
    M6_CC -f {xmlfile}

    where {xmlfile} is the xml-file produced by vex2xml.py (see above)

    Field System

    • log-into mrt-vlbi as user oper
    • Download/copy the vex file into the directory: /usr2/sched
    •  

    Troubleshooting