Created 7th Jan 2002 - cajordan
Configuring MB correlator - Offsets and Cals
- The offsets for each beam are found using the pulsar configuration ie LO at 386MHz. The program to run is called OFFSET1, and it will scan the telescope back and forth in first azimuth then elevation to produce azimuth and elevatio offsets for whichever beam the pulsar hardware is connected too.
- For beam 1 the two red sleeved coax cables should be attached to pulsar outputs 1 and 2 in the filter/attenuator racks at the LHS of th emain observing racks. The controller can run the offset program for you. noting the results.
- Repeat for all 4 beams, and calculate the offsets of the centre of the cluster of beams (average all the values!)
- In an ARTHUR window, change to the directory containing the multi configuration files by typing SD MULTI$CAL
- Edit the file LOCAL_SETTINGS.DAT so that the line
pp = -0.009 -0.054 0.159 -0.097 0.026 0.224 -0.171 -0.043 -0.052 -.354 0.0
Gives the offsets of the beam centre, followed by the offsets of beams 1,2,3,4, followed by a dummy value of 0.0 (there should be 11 values on the line)
- Restart multi/tkmulti to pick up the new offsets.
- When the correct beams offsets have been found, do a calibration run on a suitable calibrator with the correlator configured as normal ie LO at 391.25MHz and Noise diode connected.
|3C48 || 15.9||01:37:41.3|| +33:09:35.4|
|3C123 || 48.8||04:37:04.4|| +29:40:15.0|
|3C147 || 22.5||05:42:36.1|| +49:51:07.2|
|3C161 || 19.0||06:27:10.0|| -05:53:07.0|
|43.3|| 09:18:06.0|| -12:05:45.0|
|3C286 || 14.8||13:31:08.2|| +30:30:32.9|
|3C295 || 22.4||14:11:20.7|| +52:12:09.0|
| 45.2||16:51:08.3|| +04:59:26.0 |
|3C353 || 57.4||17:20:29.5|| -00:58:52.0 |
- Configure tkmulti as follows:
- Change number of cycles to 10.
- Change mode from SCAN or TRACK to CALIBRATE
- Enter coordinate type and coordinates for the chosen source
- Enter a fcc rotation of zero (this is the offset from 45 deg ref position and you should check that feed rotation is approx 45 deg. during the calibration run).
- Hit the START button
- The program will put each beam in turn onto source by setting the appropriate offsets (You should see these being set in MCDSP), and run for 10 cycles.
Only power readings are taken (no spectra) and many more than usual per 5 sec cycle. The power levels will be much lower than for a SCAN.
- Monitor each beam in turn with SPD showiing both noise diode ON and OFF for the two pols (ie sel 1,2,3,4 for beam 1). You should see a significant signal change when the beam is on source.
- You can dump the data displayed by typing file in SPD and produce a colour postscript file by typing write colour.
- When the run is finished, print the postscript files (Change to directory COR$LASER and type PRINT/Q=IRIS SPD.CPS for the final print, and
for the preceding ones.
- JB_MBD.DAT, JB_MBD.DAT;-1, JB_MBD.DAT;-2, JB_MBD.DAT;-3
are data files holding the numbers used to produce the graphs as column arranged data - which can be processed as wished!
- What is required from the data is the value in Jansky of the noise diode for each pol. of each beam. Crudely, you can do this with a ruler and pencil on the colour printouts. More accurately, you can use gnuplot, or write your own utility to find the required data.
- Finally, these numbers are enterd in the file CALPARAM.DAT in the directory multi$cal.
- multi/tkmulti must be restarted to pick up these new values.