JKT High Frequency Observing for the Perkin-Elmer + sparc IPX observing system

What is it for?

For certain observations, such as time-resolved spectroscopy, it is of maximum importance to get the maximum number of observations per hour. This note describes the software capabilities introduced to support this goal

What has been done to the software?

The Adam software on the Perkin-Elmer has had some modifications which allow certain time-consuming operations normally done for each observation to be suppressed.

So how frequently can observations be made now?

Test were done using a window at (0,0) of 100x100. Using the commands described below the elapsed time between cycles of a multrun command is 50 seconds + exposure time. Of the 50 seconds 26 is electronic readout into PE memory and 5 is the dump of that memory to a disk observation file. The remaining 19 seconds are Adam dialogues to the CCD task etc. No further reduction of elapsed time will be possible. The use of SETQUICK for a window of such small dimensions makes no appreciable difference so it need not be invoked.

Instructions for suppressing non-essential tasks

Suppression of

Suppression is done by setting Adam variables to logical TRUE or invoking an Adam procedure as described in the following table

element suppressed Adam instruction
DISPLAY on lexidata set Adam variable to logical TRUE
 Adam:> nodisp=1 
logging of observation data to night log set Adam variable to logical TRUE
 Adam:> nonight=1 
generation of image header file for use in
transfer to a sparc.
set Adam variable to logical TRUE
 Adam:> noxfer=1 
clearing of chip prior to each exposure invoke an Adam procedure
 Adam:> noclear 

Warnings on disabling elements of software

Please take time to read and understand the following. If you disable any of the elements of the observing system software to save on elapsed time between exposures you must also be aware of the consequences.

Disabling display on lexidata

There is only the danger that you will not be aware of any fault which occurs which affects the image.

Disabling logging of observation data to log

You will have no file which records the run numbers and the objects observed.

Disabling generation of image header file for use in transfer of images to the sparc

If you set the flag NOXFER to logical true (ie 1) then the complementary files IMGHDR.nnn where nnn is the 3-digit run number (001, 002, ... 999) will not be made. Without these files the images cannot be transferred using the peclient-peserver pair of programs. However, you can at any time generate these files. For example you could do observations through the night and in the morning build all the IMGHDR.nnn files and then begin the transfer process and leave the transfer running for the next few hours.

To build the IMGHDR.nnn files can be quite tedious. You have the Adam command headout which can make a single file


In this example the observers initials are GFM. To build the file for run #5 you can do either of

 Adam> headout r5 
 Adam> headout CCDGFM.005 

However you may have several hundred files to process. In this case you should make an Adam procedure file (file suffix .PRC) in your Perkin-Elmer account which has all the necessary headout commands and then invoke that procedure from Adam.

Alternatively contact the Computing Facilities Group for assistance - preferably by telephone or e-mail to helpdesk.

Disabling clearing of chip prior to each exposure

WARNING: the disabling of "clear prior to expose" is now in effect until explicitly re-enabled. A chip is normally cleared just before opening the shutter to remove any noise from light leaks or more commonly the accumulation of cosmic ray (high energy particle) traces.

If automatic clearing is disabled then the more time allowed to elapse between the end of the readout of the previous exposure and the start of the next exposure then the more noise there will be on the chip before the shutter even opens. This means that if for example the observers think for 5 minutes and then do 10 sequential exposures then the 1st may need to be discarded because it will not have been cleared of 5 minutes accumulated noise.

To force a clear at any time (eg after thinking for 5 minutes) use the command

Adam:> doclear

To re-enable automatic clear before expose use the command

Adam:> clear

Because "noclear" and some other of these suppressions could result in less information than would normally be recorded messages of the form are produced.

         "OK: xxxx not done because xxx is TRUE" 

Resetting to normal conditions

The surest and safest way of resetting all these suppression flags is
Adam:> ccdinit

Individual options may be re-enabled by setting the appropriate NOxxx Adam variables to logical FALSE ie 0.
Adam:> nodisp=0
Adam:> nonight=0
Adam:> noxfer=0
Adam:> noclear=0
In the case of NOCLEAR you may choose to invoke the procedure CLEAR.
Adam:> clear

Gary Mitchell 27-Nov-1995