Test of Position Switching with a Specified Off Position

A short (2-hr) test was made on the morning of 12 November 2001 for position switching with a specified OFF position, i.e. in contrast with a "standard OFF position" which gives tracking over the same segment of the dish as traversed by the corresponding ON. For these tests, short integrations of either 2 or 4 minutes in the ON position were used. "Standard" ON/OFFs" were alternated with ON/OFF scan pairs having specified OFF positions. The Sun rose late on into the test, and some data were taken post-sunrise. These data are also presented below as they reveal interesting results.

The observations were made with 4 x 25-MHz bandwidths centered (with frequency overlaps) at 1382.5, 1387.5, 1407.5 and 1412.5 MHz, and using 9-level recording. This range of frequencies was essentially unaffected by RFI during the measurements (except for the Sun). Doppler correction was switched off. The ON position used was chosen randomly (R.A. = 09.0hr; Dec = +16.0d), and no HI emission was seen, apart from that of our own Galaxy which lay at the high-frequency end of Subcorrelator (SBC) 4. A number of offsets were used for the "specified OFF positions", ranging over pure declination offsets of between 2*HPBW (7'12") and 2d30', and an offset in R.A. equal to that for "standard position switching", but of opposite sign.

2-Minute ON Integrations

For 9-level correlation, an rms value for (DelT/Tsys) of about 8.43x10^-4 was expected for an unsmoothed (ON-OFF)/OFF analysis, i.e. 2/sqrt(channel-width * total integration time). The values obtained from the observations (in units of 10^-4) for the central 15 MHz of each SBC were as follows:

--------------------------------------------------------------------
Offset                Poln A                   Poln B
           SBC1   SBC2   SBC3   SBC4      SBC1   SBC2   SBC3   SBC4
--------------------------------------------------------------------
Standard   8.28   8.34   8.06   8.49      8.82   8.42   8.18   8.34
--------------------------------------------------------------------
-7'12"Dec 10.39   9.93  10.37  10.06     10.13  10.80  11.11  12.30
-1d15'Dec 12.10  12.86  12.11  11.51     11.51  13.65  12.42  11.68
--------------------------------------------------------------------

4-Minute ON Integrations

Here, the theoretical rms value is about 5.96x10^-4. The values (in units of 10^-4) for pre-sunrise scans were;

--------------------------------------------------------------------
Offset                Poln A                   Poln B
           SBC1   SBC2   SBC3   SBC4      SBC1   SBC2   SBC3   SBC4
--------------------------------------------------------------------
Standard   6.46   6.07   6.01   6.33      6.04   6.12   6.12   6.33
Standard   6.28   6.25   5.87   6.14      6.22   6.09   6.43   6.41
--------------------------------------------------------------------
-1d15'Dec  9.16   8.69   9.18   9.36     11.23  10.76   9.68  10.02
-1d15'Dec  8.89   8.01  11.45  10.69      9.19   9.83  10.20  11.38
-2d30'Dec  9.53   9.73  10.29   9.84     10.49  10.58  10.20  10.18
-5mRA      8.02   8.22   8.36   8.47      9.29  10.96   8.91   8.47
--------------------------------------------------------------------

8-Minute ON Integrations

Two 8-min ON integrations were obtained by summing, a) the pair of Standard ON/OFFs, and b) the pair of scans with the OFF position:

           SBC1   SBC2   SBC3   SBC4      SBC1   SBC2   SBC3   SBC4
--------------------------------------------------------------------
Standard   6.33   6.56   6.42   6.46      7.19   6.54   6.63   6.59
Standard   6.69   6.67   6.77   6.88      7.37   7.02   6.80   7.08
--------------------------------------------------------------------
-1d15'Dec 51.19  34.91  29.89  64.27     13.54  12.84  12.83  13.65 
-1d15'Dec 24.04  24.81  25.41  19.83     29.26  42.29  25.99  22.15 
--------------------------------------------------------------------

Conclusions

Although this was a short test, and a longer, more comprehensive test may be useful, a few conclusions can be drawn from these data.

1) The "standard ON/OFF" procedure for position switching performs well, producing results close to the theoretical signal-to-noise ratio for the integration times used here.

2) Even for the rather short integrations of 2 and 4 min used here, position switching using specified OFF positions seem to result in degraded signal-to-noise ratios compared to theoretical expectations. Further observations need to be made to investigate what the situation is for different integration times (shorter, as well as longer). The effect of varying the magnitude of the offset also needs more investigation, although the data here suggest that there is little difference for 4-min integrations between declination offsets of 1.25 and 2.5 deg.

3) Sunrise effects seem to cause vastly degraded response for a specified OFF position when compared with "standard position switching".

---

For more information, contact Chris Salter

Last updated: Friday, 14-Feb-2003 11:36:48 AST