From: <Chris.Phillips_at_email.protected>
Date: Wed, 17 Nov 2004 15:33:07 +1100 (EST)

Hi all,

*************
* VERSION 2 * 17/11/2004 15:30 AEST
*************

This is an updated version of Huygens setup notes. If you are
responsible for setting up a station please read carefully. This has
been updated since the last copy, including quoted frequencies.

**************
* TIME RANGE *
**************

LBA participates is in the time range:

 14:45:00-22:30:00 UT (Wed 17 Nov)

Only the ATCA records for the entire time. Individual antenna times
are below. Only run vsib_record when the telescope is on source.
Remember that vsib_record overwrites files by default! When restarting
be careful to use different file names!

*******************
* FREQUENCY SETUP *
*******************

The basic frequency setup is:

              LSB USB LSB USB LSB USB 8 16MHz channels, 512Mbps
         RCP \_/ \_/ \_/ \_/ \_/ \_/
         LCP \_/ \_/
                                       |
         2023.99-2039.99 LSB RCP \__=> Cassini, MEX spacecraft
         2039.99-2055.99 USB RCP in this band.

         2123.99-2139.99 LSB RCP
         2139.99-2155.99 USB RCP

         2270.99-2286.99 LSB RCP
         2286.99-2302.99 USB RCP
         2270.99-2286.99 LSB LCP
         2286.99-2302.99 USB LCP

*************
* DAS SETUP *
*************

A special DAS profile is needed. See setup at bottom of file and
attached example.

This requires 2 DAS with each USB/LSB being the output from one IFP of
the DAS.

For the 4 bands the *central* frequency for the input to the DAS will be
 DAS1 IFP#1 2039.99 MHz (RCP)
 DAS1 IFP#2 2139.99 MHz (RCP)
 DAS2 IFP#1 2286.99 MHz (RCP)
 DAS2 IFP#2 2286.99 MHz (LCP)

The ATCA and Ceduna will only record the "DAS2" setup.

Please configure "DAS1" to record the low frequencies and "DAS2" to
record the high frequencies. If you cannot arrange this (ie Mopra) you
can attach the 2 headed cable into the VSIC the opposite way around.

The nominal DAS1 cable of the Huygens double header cable is the one
without the clock line (ie 11 wire pairs). The DAS2 cable is has two
sets of wires (9 data pairs and 5 clock/pps pairs).

*************
* RECORDING *
*************

This is another disk based experiment using the MRO system. Please
make sure your disks are clean or swapped as appropriate
(Mopra/Ceduna) (don't delete vt001h otherwise Craig will kill you).

As all antenna but the ATCA record in two blocks, it is best to use
dstart and vsib_record twice to do the recording. You *do not* have to
use dstart if it is easier not to (The JIVE correlator won't care),
but please do not start recording early or record after the start of
the break (limited disk space). Using dstart solves that problem.

This is a 512 Mbps recording rate at Parkes/Mopra. Please don't run
fauto etc during recording (vsib_checker -p on a couples of files
should be safe). You need to run vsib_record with the "-m 2" option to
tell it which record mode to run (16 bits of data total per sample
[8x2 bit]). You need to run with the two headed Huygens cable plugged
into both DAS! An example usage will be something like:
>> vsib_record -m 2 -t 17100s -o gg057n-Mp1-%04d < /dev/vsib

Ceduna/ATCA run with normal cabling etc as for vt001i and vsib_record
mode "-m 3" (this is the default). Record using something like:
>> vsib_record -m 3 -t 17100s -o gg057n-Cd1-%04d < /dev/vsib

Remember that vsib_record overwrites files by default! When restarting
be careful to use different file names!

***************************
* Specific Antenna notes: *
***************************

****
ATCA
****

 - Record using a single DAS for the entire run.
        UT 14:45:00-22:30:00

  - The caobs frequency should be set to 2287 MHz.
  - CATIE: function xx-yy xx+yy xx-yy xx+yy
  - DAS profile huygens_n.pro
  - Use the 4x250 GB IDE disks on the Promise adapter card (/lfs/raid0)

*****
Mopra
*****
 - Record with 2 DAS with the frequency setup above
 - Record for two periods:
            UT 14:45:00-19:30:00
        UT 20:50:00-22:30:00
 - Recording must be stopped for the "off" period
 - See dstart usage below for an example
 - Use a new set of 4x400 GB IDE disks (/data/disk2). They can be
   installed *before* vt001i to save time later.

******
Parkes
******

 - Record with 2 DAS with the frequency setup above
 - Due to elevation limitation starts late. Do not start recording until
   the beginning of the Parkes schedule. Also observes in two periods.
        UT 15:45:00-19:30:00
        UT 20:20:00-22:30:00
 - Recording must be stopped for the "off" period
 - See dstart usage below for example running
 - Use the current 4x400 GB IDE disks on the promise adapter /lfs/raid0
   (swap to the other 4x400 GB disks after the experiment).
 - The DAS profile has changed from the last Huygens test

******
Ceduna
******

 - Only record using a single DAS. The centre of the DAS input frequency
   should be 2286.99 MHz.
 - Record for two periods:
        UT 14:45:00-19:30:00
        UT 20:50:00-22:30:00
 - Recording must be stopped for the "off" period
 - See dstart usage below for example running
 - Install 2x400 GB IDE disks especially for this run (/i1). Do the
   disk swap *after* vt001i (swap back again after the experiment).
 - Remember to boot with the right kernel options and re-install
   the vsib kernel module after rebooting!
 - See setup notes below for software RAID setup

******
Hobart
******

 You are on your own here. I calculate you are recording for 6h25m, as
 you also have an 1h20m gap in the middle.
 - Recording for two periods:
        UT 14:45:00-19:30:00
        UT 20:50:00-22:30:00

************
* APPENDIX *
************

*********
DAS SETUP
*********

The DAS recording mode should give you two contiguous 16 MHz bands per
IFP (ie 4 16 MHz bands per DAS). As we are recording with the VLBA we
need to use a non-standard DAS profile. The differences from a normal
"vsop" profile are:
 - Swapped spectrum for the lower frequency of each pair. This is to
   make the data look like LSB/USB pair.
 - Using "AT" bit encoding (sign/mag) rather than "VLBA" (offset binary).
   This is only for historical purposes as we used that mod for the first
   Huygens test.

I have attached a v175b_n.pro and v175b_f.pro profile which should
work OK at Mopra (not tested!). This is based on the vsop.pro profile
(not vsop.pro should really be vsop_f.pro). To modify an existing
vsop.pro setup make a copy of the ASCII file (to v175b_f.pro say)
and change with a text editor and set:

 Y_Spectrum_Flipped: 0 0

 S2_LBits_Format: 0 0
 S2_HBits_Format: 0 0

 v175b_n.pro will be the same except it will also have:

 Swap_Sidebands: 0 0

***************
DSTART examples
***************

I suggest you create your self a little script to run these, e.g.

-------------------------
#! /bin/csh -f

etenv TZ UTC
dstart XXXXXXXXX; vsib_record -t XXXX -m 2 -o gg057n-XX-%04d < /dev/vsib
dstart XXXXXXXXX; vsib_record -t XXXX -m 2 -o gg057n-XX-%04d < /dev/vsib

----------

* ATCA

dstart 2004-11-17T14:45:00 ; vsib_record -m 3 -t 27900s -o gg057b-At-%04d < /dev/vsib

* Mopra

dstart 2004-11-17T14:44:59 ; vsib_record -m 2 -t 17100s -o gg057b-Mp1-%04d < /dev/vsib
dstart 2004-11-20T20:49:59 ; vsib_record -m 2 -t 6000s -o gg057b-Mp2-%04d < /dev/vsib

* Parkes

alias dstart ~/vsib/dstart
alias vsib_record ~/vsib/vsib_record

dstart 2004-11-17T15:44:59 ; vsib_record -m 2 -t 13500s -o gg057b-Pa1-%04d < /dev/vsib
dstart 2004-11-20T20:19:59 ; vsib_record -m 2 -t 7800s -o gg057b-Pa2-%04d < /dev/vsib

* Ceduna

alias dstart ~/vsib/dstart
alias vsib_record ~/vsib/vsib_record
dstart 2004-11-17T14:44:59 ; vsib_record -m 3 -t 17100s -o gg057b-Cd1-%04d < /dev/vsib
dstart 2004-11-20T20:49:59 ; vsib_record -m 3 -t 6000s -o gg057b-Cd2-%04d < /dev/vsib

(Note user "pb" uses a bash shell. The example script above is csh. I
have used the csh style of alias. Modify however you want.)

*********************
Ceduna softraid setup
*********************

- Follow my software "howto" but remember the exact details will be different
  (2 disks).
- Remove all current IDE disks for safety.
- Check the new disks are jumbered cable select
- Install 1 disk per IDE cable
- Format disks
- Change /etc/raidtab to just use /dev/hda1 and /dev/hdb1
- create raid disk and format filesystem
- Mount and change permissions
- Run my diskspeed program (contact me if you don't have it)
  to check the write speed is OK
- After the experiment return the 4 400 GB disks. Restore /etc/raidtab.
  I would suggest you return the disks to exactly the same position i
  cables etc as before removing them.

<div>


Received on 2004-11-17 15:33:30