The spider scan (heiles calibration scan) is the pattern that has been used to do the telescope calibration. It has 4 strips rather than the two strips of a cross. The extra two strips allow a more detailed measurement of the sidelobes as well as allowing the measurement of beam squint and squash.
A1943 (alfa galactic precursor proposal) took spider scan data in sept04 using the wapps and galfa. The goal was to get the reduction software working on the wapps and on galfa. At the time the routines did not work. This was reported and the routines were worked on.
On 30oct04 spider scans were again taken by a1943 using alfa and beam 0. The pattern completed without aborting. The ra/dec positions from this scan showed that the 3rd strip was offset from the correct path by over 1/2 a beam. The third strip missing the source caused the 2d fits to not converge.
The tcl code that drives the spider scan is /home/aoui/develop/exec/exec.d/wappmap.tcl
routine cross_wapp.
The code extends the normal length of a strip by 6 seconds to give
the telescope time to get up to speed before the datataking starts. The
code for doing this was:
offset= N*FWHM/2. .. positive offset for start of map.
azextraOff= rateAz*6
zaextraOff= rateZa*6
startOffsetAz= dirAz*(offset + azextraOff)
startOffsetZa= dirZa*(offset + zaextraOff)
The problem was occurring because rate was a signed value while
offset was always positive. If the rate was negative then (offset
+ azextraoff) would make the offset smaller rather then larger. This only
occurred for spider scans strip 3 (which had a negative za rate). All of
the other strips (cross or spiderscan) had positive rates. The fix was
to take the absolute value of the extra offsets..
On 31oct04 i changed the wappmap.tcl code to include the abs() routine.
It is marked in the code with at
<pjp001> line. Before doing this i simulated the position/offsets
in idl.
The plots show the ra/dec pattern before and after the changes (ps) (pdf) were made.
home_~philFig top: this has the 4 strips of the spider scan before the change. You can see that the green strip (3) is not going through the center of the other 3 strips. Fig bottom: This is the spider scan positions after the fix. All of the strips now pass through a common center.
processing: usr/a1943/spiderpos.pro