Focussing performance

(Last update for HET230, HET345, LABOCA, SABOCA, FLASH345)

This web page is dedicated to make a monitoring for focus observations (in all three directions). To check the behavior of the subreflector/hexapod and ambient temperature. The elevation is not considerate yet but new analysis show in the Y axis a strong component. Data since 1st June until !5th September 2011 is considered. A special case is SABOCA data because during the current year we added the map focus option for the bolometer observation, the data produced with this command also was added in this analysis. It improve the final data distribution. The form above gives the absolute focus correction selecting the variable to compare (ambient or subreflector temperature ), the FrontEnd and the three different axis. The current automatic focus correction already implemented just considerate Z for LABOCA, HET345 in the case for facility instruments.
The plots below show the relation between absolute offset and temperatures for the different receivers.There are a large data dispersion in all the plots (for all the receiver), the box start focus values help to compensate this. To use it is necessary to have a focus measure and if there's a large difference between the estimate and real value, this delta value could be added in this box and continue using the estimation.


The first plot is the absolute offset distribution, where in the case for APEX 1 the values are -0.2, 0.4, 0.25 for X, Y and Z. The second plots show the relation between the absolute offset and ambient temperature, the fit is a polynomial with order 2 in all the case. The third plots show the relation between the absolute offset and the subreflector temperatures average, the fit in this case is linear.


The information for these plots group is similar than the above, but for the plot between absolute offset and subreflector temperatures only for Z, with red line show the function than Dirk already implemented in the system.


For X and Y in CHAMP810 and SABOCA there are a large data dispersion, the plot distribution in both case are not clear. In Z for both case the data distribution look much better, but there are a big different between the current model and this fit.


---

Last update: May 24th, 2015