Six keywords are needed in order to reproduce See User and rotator angles for rotator tracking modes (all focal-stations) in the keyword world:
Refer to See KSD/46, "Keck I DCS Keyword Manual" and See KSD/106, "Keck II DCS Keyword Manual" for details of these keywords.
|
rotator angle (ROTPDEST) 1 |
||
|---|---|---|
|
ROTDEST - PARANG - sign(FOCALSTN) 2 * EL |
||
Many will be familiar with two further rotator keywords:
When the rotator is tracking, the "POSN" keywords will have values which are close to the "DEST" keywords.
See Pointing Origin Coordinate System refers to the "rotator zero point" being chosen so that at zero rotator angle the pointing origin Y axis coincides with the telescope elevation axis. The rotator zero point (ROTZERO keyword) is defined as the value which, during a full initialization, is loaded into the rotator encoder when the home switch is encountered.
ROTZERO is not convenient to use, because the effect of a change can be seen only by performing a full rotator re-initialization. A more convenient way to get the zero point right in the first place is to use the rotator calibration angle (ROTCALAN keyword) which is simply a value which is added to ROTDEST (and subtracted from ROTPOSN).
Once the zero point is correct, ROTCALAN can be absorbed into ROTZERO. A positive ROTCALAN corresponds to a negative ROTZERO adjustment for all but the Bent Cassegrain rotators (where the physical rotator angle is measured in the opposite direction on the sky), for which it is a positive ROTZERO adjustment. Once this has been done, further fine tuning may slightly change ROTCALAN. Such small changes need not be absorbed into ROTZERO (large changes must be absorbed so that guider corrections, which make use of ROTPPOSN, will be in the correct (Az,El) direction).
The rotator reference angle (ROTREFAN keyword) is equivalent to ROTCALAN in that it is (currently) added to ROTDEST and subtracted from ROTPOSN. However, it serves a completely different purpose: it's to permit ROTDEST and ROTPOSN to be measured relative to a zero that makes sense to the instrument user (e.g. a slit or an object/guide star vector) rather than relative to the YIM axis.
Currently, all ROTREFAN values are zero. We are discussing whether it would be better not automatically to apply it to ROTDEST and ROTPOSN but instead to leave that up to keyword clients. This would make the transition from the current situation (where clients apply various hard-coded instrument-specific offsets) more smooth.
Refer to See Position Angles and Rotator User Angles for exhaustive discussion of the above-mentioned instrument-specific offsets.
1. Note that these relationships do not hold precisely when EL is the telescope elevation (refer to See (Az,El) Coordinate System ). We could and should define another keyword, which would be very similar to PARANG but which would allow the relationships to be precisely correct.