Report on my stay at La Palma, 8. Apr -> 22. Apr 2006


The basic work to be done was

1. Upgrade of BIAS sw for NotCam
2. Investigate the timing in NotCam array readout
3. Writing an addedum to the BIAS sw documentation, with
   respect NotCam specific functionality, timing etc.
4. Transfer knowledge of motor controller, uProcessor 68HC11,
   software.

An email from Thomas Augusteijn to Johannes Andersen, from
14. Feb describes the requested tasks for NotCam software.

From the 10. Apr to the 18. Apr, I stayed at the telescope, the
19. Apr to 21. Apr, mostly at SLO.




The software in its present condition, with some upgrades made,
showed that the 'display' and the 'header' commands did not
work anymore. This is because a new naming of image files has
been introduced. This was not considered a problem, as images
are transferred to another PC with IRAF installed.

A number of traps, printing out the timing of the readout was
introduced in the software. This was done to analyze the differences
found in the FITS header keywords values, TM-START, TM-END and the
EXPTIME.

It was seen that the readout time of the array is around 3.6 sec,
actually I find 3.52 sec. Several tests show that TM-START is
actually is syncroneous with the sending of SINT/SSEQ commmands sent
to the array controller. The timing and readout of the images, with
the commands exp, mexp, dark, mdark, frame and dframeis from then
on determined by the controller. The start of the sequences in the
controller is started only a few uSec after the sending of the command.

The TM-END value, has a delay of 0.4 -> 1.4 sec, due to the slow
readout in the PC of the ISA PC-interface card, and also the
on-line update of the images in ds9.

As noted in an earlier email from Preben Noerregaard, the exposure in
exp and mexp is determined by the shutter, in the frames command
determined by the readout.

A description of this is in the updated NotCam BIAS document.

A new command 'clear' is introduced. This will perform a reset of
the array, unfortunately the controller insists on reading and
transmitting the arrray. This commands the takes 3.6 sec. and does
not do a readout of the actual frame. This will speedup the clearing
of a saturated array.

The FITS keyword value of EXPMODE has been changed for the mexp/mdark
commmands. It will look like this:

EXPMODE =  '%s   n/m'

where %s is the command string, f. inst. 'exp 5', n is the number in
the sequence of m multiple exposures. In the averaged image, the
keyword NCOMBINE = N was added.

The window in imexamine, which shows the radial plot was extended to
the left, allowing for more digits on the Y-axis. This was done
using the 'set lmargin nn' command to gnuplot. The margin is reset
to its original value (this means automatic) for other imexamine
commands. Pls. note that it is critical where in the sequence of
commands to gnuplot is is placed.

When doing radial plot in imexamine, the FVHM for both the WF and
HR camera are printed. This has been hard-coded in the sw.

The linear regression algorithm was carefully analyzed. A bug
was found, fortunately not having any influence on the stored
image. When fitting the line

F(x) = A + B*x

the A value was calculated wrong.

The option of only saving parts of an mexp command was not implemented.

One full day was spend with Graham Cox introducing him to the 68HC11
software for the motor controllers. Graham had allready a DOSBOX
setup installed, and succesfull compilations was done. The
necessary cross compiler, assembler etc. was installed.
Unfortunately a change in motor controller software was not tested
with the instrument.

Jens Klougart
Apr. 2006



--------------------------------------------------
Jens Klougart
Rosengårdsvej 24,
2830 Virum

Tlf. 4583 9592 / 2084 9592

email: klougart@astro.ku.dk / klougart@mail.tele.dk
---------------------------------------------------