D-line
Notes: MedOptics Dectector
Detlef Smilgies,
CHESS
INDEX
start-up
- plug in pump and chiller cables into dual extender from "dirty"
outlet: vacuum pump and chiller should start-up
- plug in powerstrip cable in to "clean" outlet and switch on
powerstrip (behind crate backdoor)
- reset water interlock (on front panel of CCD rack) - the yellow
control light should come on
- if not, check the fuses, also on the front panel (labels
for 24V and 110V are crossed over)
- switch on controller on top of crate
- the switch is close to the power plug on the controller backside
- the green LED control light on the controller frontside should
come on
- switch on computer (toggle switch to start)
- run Medoptics Windows software (under the Start button) to check
cool-down; exit when done
- start IDL control
software:
- click icon idlde
- various numbers from the IDL - CCD controller interface
are
displayed
- if all numbers are "0" or "1024", connection is lost - try
(in various stages of despair):
- restart idlde
- power off/on CCD controller and restart idlde
- power off controller, reboot PC, power on controller,
restart idlde
- cry & pray & try step 3 again
- type allezDS on
command line
- notes:
- new timing mode: CCD controller is master,
SPEC is slave (works, some sync issues remain)
- binning enabled : provide xbin and ybin on start-up; existing
correction files: 1x1 (2x2 not yet implemented)
- to stop allezDS
endless loop:
hit the letter "q"
on the keyboard (only for trouble shooting - see tricks
& traps)
cabling
- the Medoptics CCD is usually completely cabled with crate
- controller
- vacuum pump
- chiller
- shutter, controllers, and logic box
- dedicated CCD patch panel for CCD interface
- 1" shutter in front of sg slit
and Imon
monitor with
attached cable to shutter controllers on video crate
- Uniblitz shutter controllers
- one controller needed for controlling each blade of the 1/4" shutter
- two controllers needed for controlling each blade of the
large 1" shutter
- mode "normally closed"
- switch to "normally open" to test shutter
manually - switch is on front panel
- NOTE: don't forget to set mode back to "normally closed" !!
- note: the closed position means that solenoids are not
energized
- trigger on rising edge - BNC input on rear
- home-built logic box on top of MedOptics controller
- three inputs for the home-built logic box (3-fold OR)
- open by SPEC via DAC motor "shutter" and macros
opens/closes alias shopen/shclose
- open by garage door controller for intensity checks
- output to both Uniblitz controllers via cable "sh/in"
- note: the "sh/in" signal traces from the CCD patch panel to
the video patch panel to the controllers in hutch 19" rack
- garage door and controller
- check that power is turned on
- NOTE: when power is off, the shutter behaves erratically
- input "col ok" (collimator okay): voltage from I1 monitor to
show that beam is in the hutch
- input "solar ok" (solar cell okay): voltage from I2/Iprot to
show that beam stop is in place
- input "hutch door closed": from limit switch "DOOR" via main
patch panel #10
- input "garage door open/closed": from garage door to controller
(square connector)
- output controller to DC motor (Amphenol connector)
- output to open shutter from controller to shutter
logic box
- output of solar cell to current amplifier I2/Iprot (currently
not used)
- SPEC interface
- "enable": VME digital I/O to CCD "enable" input (crate
front
panel)
- "darksig": VME digital I/O to CCD "dark" input (crate front
panel)
- "ready": CCD output flag to SPEC to regain control (via V/f
and counter)
- "exposure": obsolete (drifts "HI", when not connected - as
needed) (VME digital I/O to CCD "exposure" prepared for future use)
- on control rack:
- input: CCD enable from VME digital I/O "enable"
- input: CCD controller shutter output via CCD patch panel
- output: VME counter gate input
- idea: if CCD is enabled, the CCD controller shutter
output gates the SPEC counting
- interfacing/network
- Windows NT
computer in the hutch:
- set-up/check:
- right-click icon network
neighborhood > map
network drive
- drive H:
should be mapped to \\D1\specuser
- if not:
- choose \\D1\specuser
from
- login as: specuser
- server version: d1
Samba Server 3.0.2-6.3E
- spec control
computer d1:
- all raw images are saved in default directory specuser/images
-> should be moved
eventually
to specuser/USER/raw, if
kept
- spec - IDL communication via files names.txt and times.txt on specuser/images
- spec
CCD macros (CHESS)
- ccd_on/ccd_off
alias ccdon/ccdoff:
enables/disables taking CCD images on ct, ascan, tseries,
etc.
- dark
[time] collects two dark images of duration [time] and
dezingers them
- corr
[mode] [time]: corrected image with/without
dezingering (mode=2/1)
- (current correction files are stored in
/usr/local/lib/Macros/nondist/ )
- opens/closes
alias shopen/shclose
open or close the Uniblitz shutter manually for line-up
- regular scans are supported, but without automatic correction
- for more details see Arthur Woll's help files
- low-level macros for the VME digital I/O interface are in
nondist/ (Smilgies)
shut-down
- stop IDL control
program allezDS by hitting
q
on the CCD-PC keyboard
- switch off controller
- shut down computer
- switch off shutter controllers
- NOTE: shutter solenoids can get damaged when in the open
position for prolonged time !!
- switch off powerstrip (chiller, vacuum pump)
- unplug all cables from the clean and dirty outlets
tricks
& traps
- if you only get black images from the camera:
- check whether the main beam shutter is open (control lights on
main interlock panel)
- check whether the garage door opened (check video camera)
- check whether the sample is blocking the beam - wsam will
show a typo in samth, samz
immediately
- check whether the CCD controller is tripped (see next t&t)
- if you always get the same image from the camera (even without
beam):
- when the
CCD controller trips (by itself, or water interlock failure), the IDL pogram will still produce
"images", in particular by
repeating the last successfully taken image over and over again - fix:
- stop IDL by
hitting q on the
keyboard,
- exit IDL
- restart controller by switch it off, waiting a minute, and
switching it on again
- restart IDL by
double-clicking IDL icon
- start acquisition program by typing allezDS (see start-up)
- the older IDL version required allez, then 1, then 1 again
- take a test image without beam - the old image should be gone,
and all should be dark
- take a test image with the beam in a known sample position that
produces good scattering
- (in some cases the PC may have
to be rebooted)
- the garage door controll trips repeatedly and the garage door
does not open. Fix:
- switch off garage door controller for 1 min and then turn it
back on
- there is a long delay, until the shutter opens or closes
- this is an indication that some unit providing input to the
logic box is not switched on: check
- garage door controller: switched on?
- logic box: plugged in?
- CCD controller: green control light is on?
- water interlock for CCD: control light is on?
- dark and corrected images can't be saved (error message)
- ccd root directory was not declared - fix:
- if [user] is the
user directory, type ccdroot
[user]
- this only needs to be done, when the user directory is set up.
calibration
About once a year the MedOptics detector should be recalibrated. This
involves the following corrections:
In both cases the MacCHESS x-ray generator needs to be set up in the D1
hutch (Bill Miller).
The flat field correction accounts for variation in detector
sensitivity, which arise over time and usage, as the fluorescent screen
and the CCD chip age. Typically 4 images with an average intensity of
about 10,000 counts average should be merged after background
subtraction. Using the TVX program of the Gruner group the images can
be processed to yield the new flatfield image of the detector. See TVX manual, pages 20-21.
For the distortion correction a hole mask pattern of 1 mm spacing is
imaged onto the detector. Again, TVX can be used to evaluate the
intensity pattern and generate the new correction files that will be
used for the dark and corr commands.
If binning other than 1x1 is to be used, the same steps for flatfield
and distortion have to be repeated. Typically 2x2 and 4x4 binning are
the most useful, for increasing read-out spead and/or providing more
averaging and higher S/N for diffuse scatterers, where the full
resolution is not needed.
The current correction files are kept in directory ... on the station
computer and a dated back-up should be made of each set.
p.s. Some other features are explained in A. Woll's MedOptics Notes - note that some values mentioned there are outdated, though.