NRAO Home  >  Green Bank  |  Wiki Topic:    GB > PTCS > InformationForProjectTeam > AntennaCharacterizationSpring2004 > CommissioningHistory > TPTCSPNT_070907

Changes | Index | Search | Statistics | Go

Observers: Frank and Todd

Operator: Kevin

Note: Data taken from Sep 3, 2007 through noon Sep 20, 2007 have the Antenna manager bug which causes the OBSC_EL column in the Antenna FITS file to be uncorrected for refraction.

Started script NewHalfPower? at 00:13 UT on Sep 7, 2007, weather: T=20.9, P=932mBar, DP=16.8, wind=0.5m/s

Offsetting 0.5HPBW in Elevation.

Scan numbers 2783 - 2845

2783-2787 p/f                                        avg(rms)  m/s               midpoint
2788-2792 calib    0.111/0.114 K per arcsec       Wind1    Wind2     Wind3      Azim Elev  WindDir1 Azim-WindDir
2792      way off rx_rms = 0.30"/28"  
2793      half-power-el 2007_09_07_00:26:26.fits  0.9(0.4)  0.4(0.2)  1.6(0.3)   410  63    31(17)    19
          rms = 1.55"/1.51" - rx = 1.52"/1.48", avg = 1.50"
2794-2798 p/f           2007_09_07_00:51:49.fits  
2799      half-power-el 2007_09_07_00:56:42.fits  0.44(0.20)  0.34(0.32)  1.81(0.11)   407  67    32(23)    15
          rms = 1.32"/1.30" - rx = 1.29"/1.27", avg = 1.28
2800-2804 p/f           2007_09_07_01:30:45.fits

2805-2809 calib
2809      way off rx_rms = "/"
2810      half-power-el 2007_09_07_01:39:13.fits  0.3(0.2)  0.4(0.5)  1.3(0.4)   394  73    61(18)    +333 = -27
2811-2815 p/f           2007_09_07_02:12:34.fits
2816      half-power-el 2007_09_07_02:17:34.fits  0.7(0.4)  0.5(0.2)  0.3(0.1)   374  77    238(19)   +136
2817-2821 p/f           2007_09_07_02:50:56.fits
2822      half-power-el 2007_09_07_02:55:57.fits  0.7(0.4)  0.6(0.4)  0.2(0.2)   348  77    233(13)   +115
2823-2827 p/f           2007_09_07_03:29:38.fits

2828-2832 calib   0.110/0.110 K per arcsec
2832      way off rx_rms = 0.20"/0.28" 
2833      half-power-el 2007_09_07_03:38:06.fits  0.6(0.3)  0.7(0.4)  0.3(0.2)   326  73    256(27)    +70
          rms = 1.10"/1.12" - rx = 1.09"/1.09", avg = 1.09"
2834-2838 p/f           2007_09_07_04:11:26.fits
2839      half-power-el 2007_09_07_04:16:25.fits  0.5(0.2)  0.4(0.4)  0.3(0.1)   315  69    325(34)    -10
          rms = 1.33"/1.38" - rx = 1.32"/1.35", avg = 1.34"
2840-2844 p/f           2007_09_07_04:49:44.fits
2845      half-power-el 2007_09_07_04:54:35.fits  0.5(0.4)  0.4(0.3)  0.4(0.1)   310  63    244(11)    +66
          rms = 1.14"/1.24" - rx = 1.13"/1.21", avg = 1.17"
2846-2850 p/f           2007_09_07_05:28:19.fits
----------------------

All-sky : scans 2846 - 3435

Kevin reported (at 6am) a telescope abort due to the source being at too low an elevation.

Donna reported no instances of LPC communication problems.

1:15 pm Sept 7 : Joe is doing tests of the subreflector coordinate problem.

3:55 pm  Start new half-power track script "NewHalfPower" on 2020+2942  (19:55 UT)
Starting at az=61, el=11.5, half power offset is in elevation.
(planning to run this until 5 AM tomorrow morning)

Weather is partly cloudy winds about 3-5 m/s.  Peak baselines a little sloped.
Repeated initial peak sequence (no az correction)
2nd peak elevation heuristics failed.  Changed to relaxed heuristics, start over again.
At low elevations, we really should re-balance more often than every 2 hours.
Change NewHalfPower to do only one 20-minute track after peak/focus/calibration.
The background is very variable -- not sure if this is going to work at all.

Try starting again with same source, but the new version with only one 20-minute track. Scan 3458
winds up to 6-7 m/sec
It looks hopeless.  I am going to go to all-sky pointing again until 8pm
Start pointing with scan number 3470.  Elevation is now up to 19 deg for 2020+2942
it still has problems fitting "not enough baseline present"
It looks like data wasn't being recorded for part of the scan
scan 3474 focus curve is nonsense.

Next source at higher elevation (33) baselines very flat, but data missing when
GFM tries to plot it.  gfm cant find the data???
Paul is trying to debug the problem.

Joe and Paul working on the problem.
There are a lot of NULLs in the Antenna FITS file for OBSC_AZ and OBSC_EL.

Apparently the antenna manager was off by 0.5 second from the correct time (YED).
Not sure if this is why the nulls are in the FITS file.
Joe got the Antenna computer time synchronized properly and 
asked Donna to restart the Antenna manager.  Now everything seems to be working ok.
Joe will find out when the time discrepancy started.  This could have messed up
the pointing.

All OK as of 5:40 PM (21:40 UT) scan number 3492.
All-sky pointing ok from about 5:40 PM to 8 PM.  Then Stopped and changed project code.

Worries about the bad Antenna Time.

Joe had found that the antenna manager time was 
off by 0.5 second from the true time.
The problem was fixed as of about 5:30 PM Sept 7th.

But how long had that time discrepancy existed?
Was it the cause of the corrupted Antenna FITS files?
Joe was, so far anyway, unable to track down exactly when the problem started.
It may have only been going on between 4 pm and 5:30 pm.

If there is a 0.5 second time error, that could translate to 
a position error of about 7.5 arcseconds.
To see if some of the pointing data may be shifted, 
I have listed the average LPCs from several different pointing runs,
before and after the afternoon of Sept 7, in Table 1.

Note: the first line (Sep 3-5) represents the combined data from the runs of Sept 3, 4, and 5.
Sep 7 (A) is the all-sky pointing data from September 7 before the time shift problem was discovered.
Sep 7 (B) is the all-sky pointing data from September 7 after the time shift problem was fixed.
(sigma simply means the standard deviation of the LPCs)

These numbers are plotted in Figure 1.

There does not seem to be a consistent shift of the order of 7" between the pre-Sept 7 and post-Sept 7 data, so maybe the only data that were affected were during the afternoon of Sept 7. The average LPCs can be affected by the distribution of AZ/EL, and by the temperature range during the observing run, so its hard to say if there is definitely a trend, but at least there is no obvious problem.

• Figure 1. Mean LPCs for several pointing runs.:
  

-- ToddHunter - 07 Sep 2007

Attachment:	Action:	Size:	Date:	Who:	Comment:
worry.gif	action	11598	13 Sep 2007 - 21:37	FrankGhigo	Figure 1. Mean LPCs for several pointing runs.