The calibration documentation explains
how to perform a complete calibration of a KappaCCD detector. Some of
the steps in that document need to be re-performed regularly as part
of the day to day operation of the machine. This document tries to
explain at what occasions which calibration steps should be performed.
When using denzo for integration, no recalibration is required. For
evalccd, an appropriate fluorescent sample should be measured using makesensi, and an ambi reflection should
be followed using spoton and spoton-integrate.
When a detector is first installed at a site
If the system is a system that has been delivered complete from the
factory after January 2000, all the calibration files should already
be on the Unix computer. Since the detector has been "moved", the
detector alignment must be checked (remove "detalign.vic" from the
calibration directory and run "autocal [noambi]"), but all other
information is still valid.
In any other case:
Make sure the detector ID is correct in KCCD.SET on the SERVER PC.
Make sure the DET_PHOTGAIN is correct in KCCD.SET on the SERVER PC.
From the original CD containing calibration information for the
detector, copy some files to the newly created directories:
def.cal or distor.info should be copied to
/usr/local/hklint/<detid>. Make sure the information
in that file is suitable for the version of the HKL software you
are using at your site.
sensi<detid>.kcd should be copied to /usr/local/hklint/<detid>.
The following files should be copied to /usr/local/calibration/<detid>:
distorpol.vic
badpixel.dark and badpixel.sdark
dark.kcd_frozen and sdark.kcd_frozen
responsMO.kcd
MO-incidence.coefficient [This might still be called incidence.coefficient on older
systems, but it is better to rename it]
If any of these files are not present on the CD/ROM (e.g. because
the detector was last checked in Delft before 12/1999), you will
need to determine them yourself following the instructions in the
how to calibrate a KappaCCD document.
Run the calibration procedure for "caldx", "caltheta", "calkappa" and
"makedetalign", resulting in a new file detalign.vic that
you copy to /usr/local/calibration/<detid>.
When to recalibrate what?
Another view: when does each of the calibration steps need to be performed?
makedistor and makedefcal
Only for the initial calibration. Distortion is a taper/detector
property. It can not change.
caltheta and makedetalign
Should be reperformed every time the primary beam position might have
changed: upon installation or removal of a "miracol", whenever the
tube has been changed or realigned, or when the detector has been
removed and remounted on the system.
calkappa
Should be tested after a high-impact collision to check whether the
goniostat alignment has not suffered.
makesensi and spoton
Only needs to be done when integrating using evalccd whenever a tube
with a new target material is installed for the first time
(i.e. before you measure your first data set with Cu radiation, a Cu
sensitivity calibration is required).
caldark
Some people say the dark current for a detector will never change,
other people say it changes all the time. Both groups are partly
right. In general, it will not be necessary to re-calibrate the dark
current corrections unless you are making your detector run at a
different operating temperature.
