User Tools

Site Tools


operations:documentation:lba_dbbc_config

This wiki is not maintained! Do not use this when setting up AuScope experiments!

This is an old revision of the document!


Setting up the DBBC for LBA observations

There are two steps to setting up the DBBC - the first is the working out which IFs are needed and then setting the channels appropriately. The setups below list the most common setups and how to get them working. Specific examples from previous experiments are shown where possible. These should work for 16 MHz bandwidth observations with either 4 or 8 channels. Anything different from this needs much more work…

In general though, DRUDG should have calculated the IF frequencies correctly. You can check by comparing the with the frequencies on the LBA wiki and using LOs of 1900 MHz for S-band and 8600 MHz for X-band.

See these notes on how signals are routed through a DBBC for some background info before reading further.

X-band, single polarisation

As an example, here's how V252ai has it's frequencies set up according to the LBA wiki:

V252ai frequency setup

Running drudg on the VEX file produces the following IF setup for Ke:

define  ifd01         00000000000x            
ifa=1,agc,1
ifb=4,agc,1
lo=
lo=loa,7600.00,usb,rcp,off
lo=lob,7600.00,usb,rcp,off
enddef

As the notes on how signals are routed through a DBBC describe, we only need one CoMo as there's a common input IF and all frequencies are in the band of Filter 1 (512 - 1024 MHz). So we can drop ifb and lo=lob…. The input IF is X-band RCP which in input 1 and we want automatic gain control. The input power level for Ke in CoMo A should be 38000. So ifd01 should be edited to look like this:

define  ifd01         13073035822x
ifa=1,agc,1,38000
lo=
lo=loa,7600.00,usb,rcp,off
enddef

Drudg produced a dbbc01d procedure that looked like this:

define  dbbc01d       00000000000x            
bbc02=809.00,a,16.00
bbc04=825.00,a,16.00
bbc06=841.00,b,16.00
bbc08=857.00,b,16.00
enddef

The LO frequencies are correct but we want to use BBCs 01 to 04 and CoMo A. So the revised version looks like this:

define  dbbc01d       13073035822x
bbc01=809.00,a,16.00
bbc02=825.00,a,16.00
bbc03=841.00,a,16.00
bbc04=857.00,a,16.00
enddef

Lastly, in setup01 the mark5 mode needs checking. We start with

mk5b_mode=ext,0x0000CCCC,0

However, we want to pass through all bits from the first 4 BBCs, so it needs to be changed to this:

mk5b_mode=ext,0x000000ff,1

X-band, dual polarisation

As an example, here's how V465b has it's frequencies set up according to the LBA wiki:

V465b frequency setup

Running drudg on the VEX file produces the following IF setup for Ke (note you may get some warning messages, just persist and produce the files anyway):

define  ifd01         00000000000x            
ifa=1,agc,1
ifb=4,agc,1
lo=
lo=loa,7600.00,usb,rcp,off
lo=lob,7600.00,usb,rcp,off
enddef

As the notes on how signals are routed through a DBBC describe, we need two CoMos, one for X RCP and one for X LCP. The first four BBCs will need to be one polarisation (RCP) and the second four (from CoMo B) LCP. A note to the correlator and whoever is running fringe checks should be sent. The input IF for CoMo A is X-band RCP which in input 1, and for CoMo B X LCP is on input 1, and we want automatic gain control. The input power level for Ke in CoMos A and B should be 38000 and 33000 respectively. Lastly, we might as well make the lo settings correct (although they don't actually do anything to the DBBC). So ifd01 should be edited to look like this:

define  ifd01         13075124605x            
ifa=1,agc,1,38000
ifb=1,agc,1,33000
lo=
lo=loa,7600.00,usb,rcp,off
lo=lob,7600.00,usb,lcp,off
enddef

Drudg produced a dbbc01d procedure that looked like this:

define  dbbc01d       00000000000x            
bbc02=793.00,a,16.00
bbc04=809.00,a,16.00
bbc06=825.00,b,16.00
bbc08=841.00,b,16.00
bbc01=793.00,a,16.00
bbc03=809.00,a,16.00
bbc05=825.00,b,16.00
bbc07=841.00,b,16.00
enddef

The LO frequencies are correct but we want to use BBCs 01 to 04 and CoMo A for the RCP frequencies and BBCs 05 to 07 and CoMo B for LCP. So the revised version looks like this:

define  dbbc01d       13075124605x            
bbc01=793.00,a,16.00
bbc02=809.00,a,16.00
bbc03=825.00,a,16.00
bbc04=841.00,a,16.00
bbc05=793.00,b,16.00
bbc06=809.00,b,16.00
bbc07=825.00,b,16.00
bbc08=841.00,b,16.00
enddef

Lastly, in setup01 the mark5 mode needs checking. We start with

mk5b_mode=ext,0x0000FFFF,1

This is correct in this case as we haven't changed which BBCs we're passing through, just their frequency and polarisation assignments.

S-band, single polarisation

The setup here is essentially the same as for X-band single polarisation except the IF inputs and LO frequencies will differ.

S-band, dual polarisation - 4 channels maximum

This is the configuration that is most fraught with hazard. It requires either physical recabling of the dbbc to put the correct S-band inputs into CoMos A and B, or requires the following (untested) procedure. If you use recabling, follow the X-band notes but note the IF inputs and LO frequencies will differ.

If not re-cabling, then read on…

This assumes that you will have S_band LCP through CoMo A and S-band

* Create a new configuration file in @@C:\DBBC_CONF@@ of dbbcke if needed. If the experiment shares a configuration with a previous experiment (same frequencies, polarisations, channel order) then you should be able to use that. Otherwise, read on. Copy the basic layout from @@dbbc_config_file_lba.txt@@ to your new file and then edit the frequencies. Put all of the channels at one frequency in lines 1-4 and have lines 9-12 for the other polarisation. If the schedule shows the channels as alternating between RCP and LCP, tell the LBA chat (Cormac and Chris by email) that the recorded channel ordering will be different. Start by checking the listed frequencies of each channel on the ATNF Wiki. In the corresponding entry in your experiment's config file change the frequency to (lower band edge-1900). As an example, if channel 1's lower band edge is 2209 MHz then the first line in the @@dbbc_config_file_v999a.txt@@ is then @@1 dbbc2.bit 309.00 16@@ as is the ninth line. If channel 2's lower band edge is 2225 MHz then the second and tenth lines will be @@1 dbbc2.bit 325.00 16@@. * %red%Load the configuration into the DBBC with the @@DBBC Client@@ command @@load=dbbc_config_file_v999a.txt@@ * %red%Reconfigure the DBBC with @@reconfig@@ * %red%Use the @@DBBC Client@@ to send the command @@dbbcform=geo@@ * %red%Set up the IF inputs. %red%Send these commands:
%red%@@dbbcifa=4,agc,2@@
%red%@@dbbcifc=2,agc,2@@
%red%@@pps_sync@@ * %red%Confirm that the crossover switches are in the correct position. For SLCP/SRCP, the switches should be direct, while for SRCP/SLCP you need to have the switches crossed. If you do change this, please leave a note for the next observer!. To check the status, it's easiest to do this by starting up a system monitor. Connect to ops4 with @@shh -X observer@ops4@@ and run @@monitor_system.pl@@. Select the relevant telescope & check the settings of the “Cross-over switches” * Edit the mk5b_mode line in the @@/usr2/proc/v999ake.prc@@ file. For 4 channels only, it should read as @@mk5b_mode=ext,0x00F0000F,1@@. !!!S-band, dual polarisation - 8 channels maximum This is the configuration that is most fraught with hazard. It requires either physical recabling of the dbbc to put the correct S-band inputs into CoMos A and B, or requires the following (untested) procedure. If you use recabling, follow the X-band notes but use 1900 when working out dbbc frequencies. * Create a new configuration file in @@C:\DBBC_CONF@@ of dbbcke if needed. If the experiment shares a configuration with a previous experiment (same frequencies, polarisations, channel order) then you should be able to use that one in the next step. Otherwise, read on.
Copy the basic layout from @@dbbc_config_file_lba.txt@@ to your new file and then edit the frequencies. Put all of the channels at one polarisation in lines 1-4 and have lines 9-12 for the other polarisation. If the schedule shows the channels as alternating between RCP and LCP, tell the LBA chat (Cormac and Chris by email) that the recorded channel ordering will be different. Start by checking the listed frequencies of each channel on the ATNF Wiki. In the corresponding entry in your experiment's config file change the frequency to (lower band edge-1900). Using v444b as an example, channel 1's lower band edge is 2244 MHz. The first line in the @@dbbc_config_file_v444b.txt@@ is then @@1 dbbc2.bit 344.00 16@@. The second RCP channel has a lower band edge of 2260 MHz and so the second line is @@1 dbbc2.bit 360.00 16@@. Lines 9-12 in the file are the same as lines 1-4 (as this sets these channels to matched frequencies) * %red%Load the configuration into the DBBC with the @@DBBC Client@@ command @@load=dbbc_config_file_v444b.txt@@ * %red%Reconfigure the DBBC with @@reconfig@@ * %red%Use the @@DBBC Client@@ to send the command @@dbbcform=geo@@ * %red%Set up the IF inputs. %red%Send these commands:
%red%@@dbbcifa=4,agc,2@@
%red%@@dbbcifc=2,agc,2@@
%red%@@pps_sync@@ * %red%Confirm that the crossover switches are in the correct position. For SLCP/SRCP, the switches should be direct, while SRCP/SLCP you need to have the switches crossed. If you do change this, please leave a note for the next observer!. To check the status, it's easiest to do this by starting up a system monitor. Connect to ops4 with @@shh -X observer@ops4@@ and run @@monitor_system.pl@@. Select the relevant telescope & check the settings of the “Cross-over switches” * Edit the mk5b_mode line in the @@/usr2/proc/v999ake.prc@@ file to read @@mk5b_mode=ext,0x0FF000FF,1@@

!!!S/X-band, single polarisation * Create a new configuration file in @@C:\DBBC_CONF@@ of dbbcke if needed. If the experiment shares a configuration with a previous experiment (same frequencies, polarisations, channel order) then you should be able to use that. Otherwise, read on. Copy the basic layout from @@dbbc_config_file_lba.txt@@ to your new file and then edit the frequencies. Start by checking the listed frequencies of each channel on the ATNF Wiki. In the corresponding entry in your experiment's config file change the frequency to (lower band edge-7600). Using v469a as an example, channel 1's lower band edge is 8409 MHz. The first line in the @@dbbc_config_file_v469a.txt@@ is then @@1 dbbc2.bit 809.00 16@@. Repeat for all needed channels and leave unused channels alone. * %red%Load the configuration into the DBBC with the @@DBBC Client@@ command @@load=dbbc_config_file_v469a.txt@@ * %red%Reconfigure the DBBC with @@reconfig@@ * %red%Use the @@DBBC Client@@ to send the command @@dbbcform=geo@@ * %red%Set up the IF inputs. I've assumed that the observing frequencies for X-band are between 8112 and 8624 MHz which is normal for the LBA. %red%Send these commands:
%red%@@dbbcifa=1,agc,1@@
%red%@@dbbcifc=2,agc,2@@
%red%@@pps_sync@@ * %red%For RCP, confirm that the crossover switches are in the direct position. It's easiest to do this by starting up a system monitor. Connect to ops4 with @@ssh -X observer@ops4@@ and run @@monitor_system.pl@@. Select the relevant telescope & check the settings of the “Cross-over switches”. For LCP, the crossover switches should be crossed. * Check the mk5b_mode line in the @@/usr2/proc/v469ake.prc@@ file. For 8 channels it should be @@mk5b_mode=ext,0x0FF000FF,1@@.

/home/www/auscope/opswiki/data/attic/operations/documentation/lba_dbbc_config.1364447688.txt.gz · Last modified: 2013/03/28 05:14 by Jim Lovell