Vertex Offset Pointing Metrology Issues Discussion
Last Update: JeffMangum - 09 April 2012
Discussion of Metrology Activation Issues Associated With Offset Pointing
Problems tracked at VERTEX-955: Position Anomalies Noted During Offset Pointing Runs
There are at least three related but different issues here which are causing problems with offset pointing:
Offset Pointing Motion Misinterpreted as FS
A relatively large azimuth slew can be interpreted as fast switching if the elevation slew is short, thus causing a 4 second blanking time. This can happen in offset pointing at high elevations because for a given on-the-sky-distance, the azimuth travel is much larger as compared with low elevations. With a large azimuth slew, the metrology correction can change significantly, so a 5 second exposure will be contaminated by up to 4 seconds of old and invalid metrology correction, which is resulting in failed Offset Pointing runs. Note that with offset pointing there is no sleep time in the ALMA system before an exposure is started.
Klaus confirms that if the movement of either the Az or El axis is identified as "Fast Switch" the FS blanking time of 4 seconds is applied. Klaus furthermore notes that it is certainly possible to re-define the handling of blanking times. e.g., after a slew in Az - with low acceleration - and simultaneous FS in El we would need only 2s blanking time, but due to the FS in El the FS blanking time (4s) is applied. However, an EL Fast Switch does not impact the tiltmeters as strongly as an AZ Fast Switch, because the El axis of the telescope is not a tiltmeter axis (as is Az). Thus, it would be a possibility to trigger the longer FS blanking time only in case of an Az Fast Switch
Lutz suggests the following solution:
We should implement a dependency for the metrology activation on individual axes. Furthermore I asked VA software engineering to check the filter/blanking code to see if we can gain some msec in the metrology correction.
Metrology Activation Too Large Between Two Stars in an Offset Pointing Group
If the difference in metrology correction between two sources in an offset pointing group is more than 1.5 arcseconds, the resulting jump after metrology activation will push the antenna out of the tolerance window and will cause the exposure to fail. Changing the sleep time to 2 seconds in this case would not help because in offset pointing, the exposure starts before 2 seconds.
Changes in AN0/AW0 Cause Increase in Metrology Correction
The difference in metrology correction between two different azimuths is mainly caused by changes to ANo and AWo. That is to say, immediately after those values are updated, there should be no problems, however due to changes in temperature, solar illumination, and other causes we know and expect that ANo and AWo do not remain stable. So, the metrology system must obviously allow for these variations (If they did not change, we would not need the tiltmeters!).