AuScope VLBI Operations Wiki

User Tools

Site Tools

You are here: start » operations » documentation.ivs
Trace:
operations:documentation.ivs

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

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision 		Previous revision
operations:documentation.ivs [2010/12/08 01:52]
Jim Lovell 		operations:documentation.ivs [2013/10/03 05:22] (current)
Imogen Jones [Setting Up]
Line 3: Line 3:
  
   * [[operations:​documentation.ivs.drudg|Preparing the schedule and procedure files]]   * [[operations:​documentation.ivs.drudg|Preparing the schedule and procedure files]]
-  * DBBC configuration +  * [[operations:​documentation.DBBCSetupIVS|DBBC configuration]]
-  * +
   * [[operations:​documentation.ivs.mark5|Mark5B recorder setup]]   * [[operations:​documentation.ivs.mark5|Mark5B recorder setup]]
   * [[operations:​documentation.ivs.SchedStart|Running the schedule]]   * [[operations:​documentation.ivs.SchedStart|Running the schedule]]
- 
-**Drudg** 
- 
-The .skd file was drudged on pcfsya. The rack and recorder hardware was first selected (option 11) and set to "Mark5 rack" "​Mark5B recorder"​ "No second recorder"​ and "Start with 1" (options 16 14 1 1). Then the procedures and snap file were made, and a summary printed (options 12, 3 and 5 in sequence). Printing from drudg does not seem to work at present so the summary was printed using the command "​enscript -r -2 /​tmp/​DR.tmp"​ from the terminal. 
- 
-After drudging the schedule, the procedure file was copied to /​usr2/​proc/​. It then had to be edited to remove procedures using non-existent equipment and to support the Mark5B. The ifdsx and vcsx8 procedures were removed in entirety (as the IF power levels and VC settings are set through the DBBC in our system. It may be worth keeping a copy of the vcsx8 settings to confirm the DBBC frequencies required) and the calls to these functions were removed from the setupsx procedure. The pcal functions were also removed from the proc file as we do not have a computer controlled phasecal. The vsi commands could be removed but do not seem to be causing any serious problems or excessive errors in the logs. The preob and midob procedures were edited to include some mk5 and site specific functions. One problem encountered was difficult in setting up thee new procedures by editing the proc file directly. The best way of inserting/​editing procedures in a proc file is to load it into the fs and then run pfmed from the terminal. To create or edit a procedure, type "​edit"​ and then the name of the procedure. The final procure file is included below. Don't try to paste this text into a proc file as the Wiki formatting causes strange problems for the interpreter. 
- 
-define ​ proc_library ​ 10252181423x\\ 
-" r4447      hobart12 ​ hb\\ 
-" drudg version 2008oct08 compiled under fs  9.10.04\\ 
-"< mark5    rack >< mark5b ​  ​recorder 1>\\ 
-enddef\\ 
-define ​ exper_initi ​  ​10252181420x\\ 
-proc_library\\ 
-sched_initi\\ 
-mk5=dts_id?​\\ 
-mk5=os_rev?​\\ 
-mk5=ss_rev?​\\ 
-mk5=status?​\\ 
-enddef\\ 
-define ​ setupsx ​      ​10252181440x\\ 
-tpicd=stop\\ 
-mk5b_mode=ext,​0x55555555,​2\\ 
-mk5b_mode\\ 
-bank_check\\ 
-tpicd\\ 
-enddef\\ 
-define ​ preob         ​10252182950\\ 
-onsource\\ 
-systemp12\\ 
-mk5=dot?\\ 
-clkoff\\ 
-enddef\\ 
-define ​ midob         ​10252183008\\ 
-onsource\\ 
-clkoff\\ 
-sy = run setcl adapt &\\ 
-mk5=dot?\\ 
-enddef\\ 
-define ​ systemp12 ​    ​10252182950x\\ 
-sy=/​home/​oper/​systemp12.sh\\ 
-enddef\\ 
- 
-**DBBC** 
- 
-After editing the proc file, it is necessary to set up the DBBC and Mark5B. The C:​\DBBC_CONF\dbbc_config_file.txt has been edited to set the frequencies correctly for R1/R4 type experiments. For other experiments,​ the DBBC frequencies can be calculated either from the sky frequencies or from the original values of the VCs in the proc file. The sky frequencies are given in the skd file. To convert the VC frequency settings to DBBC frequencies you need to add 480 MHZ for the first 4, add 980.1 for dbbc05-dbbc08,​ add 120 for dbbc09-dbbc14 and set dbbc15 and dbbc16 to something innocuous (say, 320 MHz) as they'​re not being recorded. Running C:​\DBBC_CONF\DBBC Control.190710.exe and answering yes to configure the system should load the frequencies correctly. One the configuration is complete, you need to set up the IF system. The commands to do this are  
- 
-dbbcifa=1,​agc,​1\\ 
-dbbcifb=4,​agc,​4\\ 
-dbbcifc=2,​agc,​2\\ 
-dbbcifd=2,​agc,​2 
- 
-It appears that CoMo 3 is reporting low total power levels. It may be that there is a poor connection on one of the SMAs...The target values for the total power readings is 56000. this can be checked by the command dbbcif[abcd]. If the levels are too high or low, then adjust the attenuation on the IF box. 
  
 **Mark5B** **Mark5B**
  
-As oper@mk5hb, first check if the dimino program is running (ps -ef | grep dim). If not, start it with "​dimino -m0&"​. While you**Drudg**+As oper@mk5ya, first check if the dimino program is running (ps -ef | grep dim). If not, start it with "​dimino -m0&"​. While your at the prompt, check that the ntp time is correct by running "ntpq -np" (for a non-interactive session). Alternatively,​ run ''​ntpq''​ and then issue the "​peers"​ command. The offset and delay to tac32ho should be small (less than ~5 milliseconds). If it is not, then you will need to resynch the machine to the NTP server. As root, issue the commands
  
-The .skd file was drudged on pcfsya. The rack and recorder hardware was first selected (option 11) and set to "Mark5 rack" "​Mark5B recorder"​ "No second recorder"​ and "Start with 1" (options 16 14 1 1). Then the procedures and snap file were made, and a summary printed (options 12, 3 and 5 in sequence). Printing from drudg does not seem to work at present so the summary was printed using the command "​enscript -r -2 /tmp/DR.tmp"​ from the terminal.+''/​etc/​init.d/ntp stop''​\\ 
 +''​ntpdate cnsshb''​\\ 
 +''​/etc/init.d/ntp stop''​
  
-After drudging the schedule, the procedure file was copied to /​usr2/​proc/​. It then had to be edited to remove procedures using non-existent equipment and to support the Mark5B. The ifdsx and vcsx8 procedures were removed in entirety (as the IF power levels and VC settings are set through the DBBC in our system. It may be worth keeping a copy of the vcsx8 settings to confirm the DBBC frequencies required) and the calls to these functions were removed from the setupsx procedure. The pcal functions were also removed from the proc file as we do not have a computer controlled phasecal. The vsi commands could be removed but do not seem to be causing any serious problems or excessive errors in the logs. The preob and midob procedures were edited to include some mk5 and site specific functions. One problem encountered was difficult in setting up thee new procedures by editing the proc file directly. The best way of inserting/​editing procedures in a proc file is to load it into the fs and then run pfmed from the terminal. To create or edit a procedure, type "​edit"​ and then the name of the procedure. The final procure file is included below. Don't try to paste this text into a proc file as the Wiki formatting causes strange problems for the interpreter. 
  
-define ​ proc_library ​ 10252181423x\\ +After entering these commands, check the ntp status again. Once the NTP time is ok, you will need to run "fmset" on pcfshb, set the time (with "​."​) and then sync the formatter ("​s"​, followed by "​y"​). Exit from fmset as soon as you have finished.
-" ​r4447      hobart12 ​ hb\\ +
-" ​drudg version 2008oct08 compiled under fs  9.10.04\\ +
-"< mark5    rack >< mark5b ​  ​recorder 1>\\ +
-enddef\\ +
-define ​ exper_initi ​  ​10252181420x\\ +
-proc_library\\ +
-sched_initi\\ +
-mk5=dts_id?​\\ +
-mk5=os_rev?​\\ +
-mk5=ss_rev?​\\ +
-mk5=status?​\\ +
-enddef\\ +
-define ​ setupsx ​      ​10252181440x\\ +
-tpicd=stop\\ +
-mk5b_mode=ext,0x55555555,​2\\ +
-mk5b_mode\\ +
-bank_check\\ +
-tpicd\\ +
-enddef\\ +
-define ​ preob         ​10252182950\\ +
-onsource\\ +
-systemp12\\ +
-mk5=dot?​\\ +
-clkoff\\ +
-enddef\\ +
-define ​ midob         ​10252183008\\ +
-onsource\\ +
-clkoff\\ +
-sy = run setcl adapt &\\ +
-mk5=dot?​\\ +
-enddef\\ +
-define ​ systemp12 ​    ​10252182950x\\ +
-sy=/​home/​oper/​systemp12.sh\\ +
-enddef\\+
  
-**DBBC** +It's worthwhile making a test recording using the mark5b. To do this you can either use tstdimino on mk5ya or the fs. The instructions below are for the fs but to use them in tstdimino, omit the leading "​mk5="​.
- +
-After editing the proc file, it is necessary to set up the DBBC and Mark5B. The C:​\DBBC_CONF\dbbc_config_file.txt has been edited to set the frequencies correctly for R1/R4 type experiments. For other experiments,​ the DBBC frequencies can be calculated either from the sky frequencies or from the original values of the VCs in the proc file. The sky frequencies are given in the skd file. To convert the VC frequency settings to DBBC frequencies you need to add 480 MHZ for the first 4, add 980.1 for dbbc05-dbbc08,​ add 120 for dbbc09-dbbc14 and set dbbc15 and dbbc16 to something innocuous (say, 320 MHz) as they'​re not being recorded. Running C:​\DBBC_CONF\DBBC Control.190710.exe and answering yes to configure the system should load the frequencies correctly. One the configuration is complete, you need to set up the IF system. The commands to do this are  +
- +
-dbbcifa=1,​agc,​1\\ +
-dbbcifb=4,​agc,​4\\ +
-dbbcifc=2,​agc,​2\\ +
-dbbcifd=2,​agc,​2 +
- +
-It appears that CoMo 3 is reporting low total power levels. It may be that there is a poor connection on one of the SMAs...The target values for the total power readings is 56000. this can be checked by the command dbbcif[abcd]. If the levels are too high or low, then adjust the attenuation on the IF box. +
- +
-**Mark5B** +
- +
-As oper@mk5hb, first check if the dimino program is running (ps -ef | grep dim). If not, start it with "​dimino -m0&"​. While your at the prompt, check that the ntp time is correct by running "​ntpq",​ and then issuing the "​peers"​ command. The offset and delay to tac32ho should be small (less than ~5 milliseconds). If it is not, then the best thing to do is to run "​fmset"​ on pcfsya, set the time (with "​."​) and then sync the formatter ("​s",​ followed by "​y"​). Exit from fmset as soon as you have finished. +
- +
-It's worthwhile making a test recording using the mark5b. To do this you can either use tstdimino on mk5hb or the fs. The instructions below are for the fs but to use them in tstdimino, omit the leading "​mk5="​.+
  
 mk5=mode=ext:​0x55555555:​2 ​ (For VSI recording, 1-bit sampling, 8 MHz BW) mk5=mode=ext:​0x55555555:​2 ​ (For VSI recording, 1-bit sampling, 8 MHz BW)
Line 133: Line 31:
 mk5=protect=off\\ mk5=protect=off\\
 mk5=reset=erase mk5=reset=erase
 +
 +Make sure there are no other experiments on the module or you will erase those too with this command!!
 +
  
 **Running the Schedule** **Running the Schedule**
Line 141: Line 42:
  
 //​Clocks//​\\ //​Clocks//​\\
-  * Run "/​usr/​bin/​xterm -name monit2 -e /​usr2/​fs/​bin/​monit2"​ on pcfsya ​to check pcfs time.\\+  * Run "/​usr/​bin/​xterm -name monit2 -e /​usr2/​fs/​bin/​monit2"​ on pcfshb ​to check pcfs time.\\
   * Check SNTP Server OK in the HMI\\   * Check SNTP Server OK in the HMI\\
   * Run clkoff to check formatter-GPS time offset   * Run clkoff to check formatter-GPS time offset
Line 150: Line 51:
   * disk_pos should report three values - the current number of btyes recorded, bytes at start of previous scan and bytes at start of current scan. If not currently recording, the first and third values should agree.\\   * disk_pos should report three values - the current number of btyes recorded, bytes at start of previous scan and bytes at start of current scan. If not currently recording, the first and third values should agree.\\
   * decoder time - Check time offset in the formatter with "​mk5=dot?"​ and check that it reports a small offset as the final value, than syncerr_eq_0 and that FHG_on or FHG_off depending on whether it is currently recording or not.\\   * decoder time - Check time offset in the formatter with "​mk5=dot?"​ and check that it reports a small offset as the final value, than syncerr_eq_0 and that FHG_on or FHG_off depending on whether it is currently recording or not.\\
-  * Check mode with "​mk5=mode?"​. The result should be "/​mk5/​!mode?​ 0 : ext : 0x55555555 : 2 : 2 ;"+  * Check mode with "​mk5=mode?"​. The result should be "/​mk5/​!mode?​ 0 : ext : 0x55555555 : 2 : 2 "; (where the last number is either 1 or 2, the number of disk modules loaded
  
  
Line 157: Line 58:
 End of experiment - The "​source=stow"​ command does not currently work so please stow the dish using the HMI interface and the "​Stow"​ button. End of experiment - The "​source=stow"​ command does not currently work so please stow the dish using the HMI interface and the "​Stow"​ button.
  
-To run the status monitor, enter this command at the pcfsya ​prompt ​+To run the status monitor, enter this command at the pcfshb ​prompt ​
  
 /​usr/​bin/​xterm -name monit2 -e /​usr2/​fs/​bin/​monit2 /​usr/​bin/​xterm -name monit2 -e /​usr2/​fs/​bin/​monit2
Line 163: Line 64:
 **Common problems** **Common problems**
  
-  * //clkoff reading is drifting or far from the maser-GPS offset// - This usually is caused by the DBBC. First, go around the back of rack 14 and move the cable from the DOTMON output of the Mark5B to the "1 PPS Mon" output of the DBBC (left hand side, sixth SMA from the top). If the same offset is seen on the counter, then the problem in in the DBBC. A temporary fix can be achieved with ''​pps_sync''​ in DBBC Control but this did not reliably fix the problem on 13/10/10. Instead, try reconfiguring the DBBC with ''​reconf''​ - this will take ~two minutes in total and you will need to re-issue the ''​dbbcifa=...''​ commands, and resynch the mark5B with ''​fmset''​.r at the prompt, check that the ntp time is correct by running "​ntpq",​ and then issuing the "​peers"​ command. The offset and delay to tac32ho should be small (less than ~5 milliseconds). If it is not, then the best thing to do is to run "​fmset"​ on pcfsya, set the time (with "​."​) and then sync the formatter ("​s",​ followed by "​y"​). Exit from fmset as soon as you have finished.+  * //clkoff reading is drifting or far from the maser-GPS offset// - This usually is caused by the DBBC. First, go around the back of rack 14 and move the cable from the DOTMON output of the Mark5B to the "1 PPS Mon" output of the DBBC (left hand side, sixth SMA from the top). If the same offset is seen on the counter, then the problem in in the DBBC. A temporary fix can be achieved with ''​pps_sync''​ in DBBC Control but this did not reliably fix the problem on 13/10/10. Instead, try reconfiguring the DBBC with ''​reconf''​ - this will take ~two minutes in total and you will need to re-issue the ''​dbbcifa=...''​ commands, and resynch the mark5B with ''​fmset''​.r at the prompt, check that the ntp time is correct by running "​ntpq",​ and then issuing the "​peers"​ command. The offset and delay to tac32ho should be small (less than ~5 milliseconds). If it is not, then the best thing to do is to run "​fmset"​ on pcfshb, set the time (with "​."​) and then sync the formatter ("​s",​ followed by "​y"​). Exit from fmset as soon as you have finished.
  
-It's worthwhile making a test recording using the mark5b. To do this you can either use tstdimino on mk5hb or the fs. The instructions below are for the fs but to use them in tstdimino, omit the leading "​mk5="​.+It's worthwhile making a test recording using the mark5b. To do this you can either use tstdimino on mk5ya or the fs. The instructions below are for the fs but to use them in tstdimino, omit the leading "​mk5="​.
  
 mk5=mode=ext:​0x55555555:​2 ​ (For VSI recording, 1-bit sampling, 8 MHz BW) mk5=mode=ext:​0x55555555:​2 ​ (For VSI recording, 1-bit sampling, 8 MHz BW)
Line 186: Line 87:
  
 //​Clocks//​\\ //​Clocks//​\\
-  * Run "/​usr/​bin/​xterm -name monit2 -e /​usr2/​fs/​bin/​monit2"​ on pcfsya ​to check pcfs time.\\+  * Run "/​usr/​bin/​xterm -name monit2 -e /​usr2/​fs/​bin/​monit2"​ on pcfshb ​to check pcfs time.\\
   * Check SNTP Server OK in the HMI\\   * Check SNTP Server OK in the HMI\\
   * Run clkoff to check formatter-GPS time offset   * Run clkoff to check formatter-GPS time offset
Line 202: Line 103:
 End of experiment - The "​source=stow"​ command does not currently work so please stow the dish using the HMI interface and the "​Stow"​ button. End of experiment - The "​source=stow"​ command does not currently work so please stow the dish using the HMI interface and the "​Stow"​ button.
  
-To run the status monitor, enter this command at the pcfsya ​prompt ​+To run the status monitor, enter this command at the pcfshb ​prompt ​
  
 /​usr/​bin/​xterm -name monit2 -e /​usr2/​fs/​bin/​monit2 /​usr/​bin/​xterm -name monit2 -e /​usr2/​fs/​bin/​monit2
/home/www/auscope/opswiki/data/attic/operations/documentation.ivs.1291773166.txt.gz · Last modified: 2011/10/26 06:37 (external edit)

Page Tools

Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki