Mini-Diff solutions
Beam Stop/Collimator badly aligned
By far the most common problem is the beamstop/collimator alignment. The beamstops are very small now (~500um(horizontal)x~300um(vertical) and can easily be knocked by careless or impatient users. If you knock the beamstop/collimator you will hopefully be able to recover by just pulling the G-ring assembly towards yourself on it's supportin mount or if that alone doesn't work by adjusting the 'bstopz' and 'bstopy' motor in exes and taking another image. Remember to set 'bstopz' motor to 0 in spec (EH2) before taking another image as the data collection procedure will move the beamstop to zero first! For more information on aligning the beamstop see the backstop/collimator page.
Cannot move the light in
There was a general problem with the Wago device servers which control the backlight. This may have been fixed as a new device server has been installed. The fix in case of the problem is to:
-
telnet device server computer (see below for user/passwd)
Beamline Computer 14-1 lid147 14-2 lid145 14-4 lid144 23-1 lid232 23-2 lid233 29 lid292 -
Kill the server by typing bliss_dserver stop Wagods
-
Restart the server by typing bliss_dserver start Wagods
-
Type reconfig in spec session
-
Restart the exes_exp and exes_hutch program
Exes won't start or you can't centre, move sample, etc
Occasionally you can no longer control the phi from exes_exp or exes_hutch and when you kill the exes jobs you can no longer restart them. One way to kill exes is to type CTRL^\ (control backslash) in the shell exes is running from. To over come this problem on the beamline control PC type ps -ef | grep exes. Kill all jobs you see except for the one which has HardwareRepositoryServer in it's job name. To kill jobs type kill "job number". Try to restart exes_exp or exes_hutch now.
If it still doesn't work:
Phi Control lost
The phi axis doesn't spin during the data collection (with an error in ProDC or spec session) or the crystal does not turn when you do your sample centering.
1. Try to type in EH2 spec session, phi_init. This command will send the phi to search for the home position (phi should spin then, check with the camera)
2. If the problem persists (for local contact only):
-
telnet device server computer
(see below for user/passwd)Beamline Computer 14-1 lid147 14-2 lid145 14-4 lid144 23-1 lid232 23-2 lid233 29 lid292? -
Kill the galil server by typing bliss_dserver stop galil
-
Restart the galil server by typing bliss_dserver start galil
-
Type phi_init, reconfig and resetup in the spec session
3. If this does not solve the problem you may need to take drastic action like turning the galil controller off/on as there is still functional problem with the controller which is being investigated. The galil controller is located at the in one of blue racks associated with each beamline, see picture below. Before you do anything, check the light status, in normal operation the lights are all green.
The bottom light should always be green, if it is red then check the air pressure gauge on the back of the mini-diff in the experimental hutch, see picture later.If the air pressure light is green then try switching the controller off/on. The controller can only be turned off from the back. So you need to go round the back and switch the controller off and on from behind (large orange switch). PLEASE TAKE CARE WHEN DOING THIS. Then use the commands described above (2) for restarting the software.
Please note in the logbook if any of the lights are red.
If the air pressure light is red and the air pressure gauge at the back of the mini-diff (side nearest the slits) is below 10bar then the controller will stop as the spindle axis can be damaged at lower air pressures. If this is the case, the booster unit supplying high-pressure air may have failed. The local contact will know where the booster is and can check this. If it is the booster, instructions are stuck on each booster showing how to bypass it and use an alternative booster. Please make sure you inform the technical staff that a booster has failed.
Only see black image instead of sample in exes
You only see black in the sample window for crystal centering. You should always see somthing in the sample window. The first thing to do is zoom out, move the beamstop out and put the backlight in. Still nothing then perhaps try recalibrating the "zoom", see below.
Still nothing, then try restarting the video device server:
-
telnet device server computer
(see below for user/passwd)Beamline Computer 14-1 lid147 14-2 lid145 14-4 lid144 23-1 lid232 23-2 lid233 29 lid292? -
Kill the falcon server by typing bliss_dserver stop falconDS
-
Restart the falcon server by typing bliss_dserver start falconDS
-
Restart exes_exp and exes_hutch applications
It is possible, but rare, that the backlight has been badly knocked. So if you still don't see anything it's a good idea to get a torch or use the extra fibre-optic cord (should find it around the light source) to shine a light directly at the camera. If you still don't see anything its likely to be a blown fuse on the mini-diff. These have to be changed by an expert (technican or local contact) and Maatel are looking into the problem.
To change the fuse you first need to turn off the power, you'll find the power switch at the back. Then unscrew the hook at the top of the mini-diff and remove the cover. You will then see three fuses at the side - picture below:
You need to unscrew the knob and change the fuse. Then replace the cover, power on the mini-diff and restart the device servers:
-
telnet device server computer
(see below for user/passwd)Beamline Computer 14-1 lid147 14-2 lid145 14-4 lid144 23-1 lid232 23-2 lid233 29 lid292? -
Kill and restart the falcon server by typing bliss_dserver restart falconDS
-
Restart the exes_exp and exes_hutch applications
Recalibrating the zoom
Occasionally the zoom motor will lose steps and you cannot zoom out enough to see your crystal. In the lowest zoom, (1), you should be able to see almost all the YAG screen. If this is not the case then simply recalibrate the zoom by:
-
typing zoom_init in the SPEC session
-
exes_exp and exes_hutch applications
Sample centering does not work
We have had problems with the sample centering that appears to have been resolved but I have included it here in the Trouble Shooting section in case it reappears. Apparently the sample centering needs the exact dial positions of the phi axis to work. So if you experience sample centering problems, simply type phi_init in the spec session, followed by a reconfig, then restart EXES_EXP and EXES_HUTCH and try again. Another possible problem is the sampy motor is driven by the software but it doesn't actually move and centering will fail. If no technician is around to fix this put a standard goniometer on and realign by hand. Get the sample on a rotation axis and then use phi vertical in exes to bring to beam position.
Additionally if the correction to be applied to get a proper crystal centring is too large and may lead the motor above the limits, the correction won't be applied. So please check first if the magnetic base is well mounted on the phi axis or if the pin is properly mounted on its base (if it is bent, we may be not able to apply the proper correction as the motors have only few mm range of movement).
If you feel it could be due to motor problem (position lost, etc...), you can reinitialize the motor positions by running "minidiff_init sampy" and "minidiff_init sampx" in spec(exp) session.
Device server login is blissadm