|
|
|
____________________
|
|
|
|
|
Continuous Calibration
  
Next: Switching Modes
Up: Calibration
Previous: Calibration
This new option uses the reference position of every cycle as the SKY in the
calibration; it calculates and uses a new  every cycle.
This has several advantages:
- Calibration using the CAL routine can be done less frequently than before,
as the sky variations are taken into account automatically. Probably once
every 30-60 mins between CALs will be fine.
- The CALs need NOT be done off-source (as only the HOT and COLD data is
actually used).
- General calibration should be more reliable - particularly in variable
sky conditions and/or at high frequencies.
- The display is updated every cycle with the current running mean spectrum;
this allows you to check baselines and data continuously, and not have
to wait until the end of the complete sample. If you do many cycles, and
see on the screen that the signal:noise is high enough before the end of
scan, the integration can be halted (at the end of next cycle), and the data
will be stored OK.
The disadvantage is slightly higher overheads (by a few %), but the less
frequent CAL's will compensate for this. For very short cycle times
(less than 15 seconds), the display will not be updated, as the
overheads will become
significant. One problem sometimes found with continuous calibration is
that, if the total power levels from the receiver are drifting
(due to problems with instrumental stability), then
line calibration problems can sometime be seen due to incorrect values
of  being calculated and applied. This has occasionally
been noted with RxA2 around 265GHz; see
section 12 for more details.
The old discrete calibration method can still be used by
answering NO to the "Continuous Calibration?" question when the DAS is loaded.
Henry Matthews
Wed May 1 15:19:04 HST 1996
|
|
