ID13 Troubleshooting

This section is designed to provide a first point of reference if any problems are encountered with the beamline. This is by no means an exhaustive list but may enable some of the more common problems to be identified and fixed.

Below are outlined some descriptions of common problems. If you see a description that matches your problem then click on it to learn more

Problems

Detector problems

The MARCCD detector software is not responding and the detector is permanently in integration mode

The MARCCD detector software seems ready and awaiting input but does not change the display after a SPEC command

The MARCCD detector software is running correctly but no files are being saved to the hardisk

The image on the MARCCD is showing red pixels near to the beam centre

The image on the MARCCD is showing coloured horizontal or vertical lines across the detector

SPEC problems

SPEC is not responding and appears to have crashed

Experimental problems

There is no intensity

There is low intensity

The hutch interlock cannot be opened

Solutions

Detector solutions

The MARCCD detector software is not responding and the detector is permanently in integration mode

PROBLEM: This is a common problem and is often caused by a user break in the wrong cycle of the detector readout (^c) however it can also for no apparent reason. When this problem occurs, the MARCCD software becomes 'stuck' in the integration cycle and does not respond to further SPEC commands.

SOLUTION: By resetting the MARCCD software the problem can be fixed. Within SPEC enter the command 'MARRESET' to cause the MAR software to be re-loaded. Make sure that after you have done this, you use the 'ccdnewfile' command to save the data and that ccd saving is set to on.

The MARCCD detector software seems ready and awaiting input but does not change the display after a SPEC command

PROBLEM: Although the MARCCD software seems to be functioning correctly (i.e. does not permanently state 'integrating' for example), the screen does not update even after a scan or count SPEC command.

SOLUTION: The most likely cause of this problem is that the MARCCD is turned off. Either turn it on in the 'blmenu' or use the 'maron' command from SPEC.

The MARCCD detector software is running correctly but no files are being saved to the hardisk

PROBLEM: Although the MARCCD detector software is responding to SPEC, either the filename string is always blank or the diffraction files are not appearing on the computer hardisk.

SOLUTION: There are two possible causes of this problem. The first is that CCD saving is not turned on which prevents the diffraction files being saved. To correct this select the option in the 'blmenu' within SPEC. The second possible cause is that the hardisk of the computer where the files are being saved is full. A quick solution to this is to copy some of the files from the hardisk to the NICE server (making sure the file dates are recorded) and then deleting them from the hardisk. The alternative solution is to specify a new disk for image saving (note that MARCCD1 and MARCCD2 directories are on different drives so simply switching between them may be sufficient). This can be done using the 'ccdnewfile' command.

The image on the MARCCD is showing red pixels near to the beam centre

PROBLEM: Some of the pixels near to the beam centre position are shown as red on the MARCCD display.

SOLUTION: This a sign of over-exposure of the MARCCD. If there are only a few pixels and they are not in a region of interest then it is usual better to use it as it is and no damage to the MAR should occur. If correcting the problem is necessary, fine adjustments of the beamstop position can normally fix the problem. Use the count-tweak command 'ctw' to adjust the beamstop position until the problem is removed:

ctw sbsy 0.01 5 This is a 'y' axis movement in 10 micron steps using a 5 second scan time
ctw sbsz 0.01 5 This is a 'z' axis movement in 10 micron steps using a 5 second scan time

The image on the MARCCD is showing coloured horizontal or vertical lines across the detector

PROBLEM: Coloured lines appear on the detector screen across the image. These are usually red or white and can run horizontally or vertically.

SOLUTION: This problem indicates a serious over-exposure of the MAR detector. Continuing operation may lead to permanent damage. First visually confirm that the beamstop is still in position (i.e. it has not been accidentally broken or knocked). If it is damaged then please inform your local contact (before 11pm) or the floor operator (after 11pm). If there is no apparent damage or position change to the beamstop then try a very short scan (ct 1) to see if the problem persists. If using a short scan the problem is reduced, it is likely that the beamstop needs repositioning. Follow the instructions given above but using a shorter scan time. If this does not solve the problem then inform the local contact (before 11pm).

SPEC solutions

SPEC is not responding and appears to have crashed

PROBLEM: This mostly occurs during a scan and involves SPEC crashing or hanging. Under these conditions it is not possible to enter any new SPEC commands and typing '^c' will not solve the problem.

SOLUTION: Before continuing, ensure that the problem cannot be resolved by issuing a break command (^c).

If SPEC was scanning, restarting it will keep the motors at their current positions. Therefore it is advisable before restarting SPEC to note down the current position of the motor being scanned (if a scan was taking place). This can then be used to return the motor to the initial position after the restart (which the scan command does automatically at the end of the scan). Exit SPEC using the '^z' command which will return the computer to the UNIX prompt. Then try restarting SPEC (using the EH2 command for the second hutch or EH1 for the first hutch). It is possible that this will return an error stating that SPEC is already running. If this is the case then it is necessary to manually terminate the existing SPEC by typing the following commands:

!ps            This lists current processes and gives their process numbers
kill xxxx     Where xxxx is the process number for the old SPEC process
!ps            Use this again to check that the process has been terminated

If the old SPEC process has been removed from the system, SPEC can be restarted although it should be noted that all limits will be reset to their previous values (the last time the '!ps' command was used). If the old SPEC process is still running, repeat the above steps but using the 'kill -9 xxxx' command instead of 'kill xxxx' (where xxxx is the process number). SPEC can then be restarted.

As a final point, it should be noted that after restarting SPEC it will also be necessary to run the 'ccdnewfile' and 'xxxx' commands in SPEC. These set the paths for diffraction pattern saving and SPEC log file saving.

Experimental solutions

There is no intensity

PROBLEM: There is no intensity on the detector (and/or only background intensity registered on the 'counter 6' channel if the ionisation chamber is installed) following a scan even though the display of the MARCCD is updating.

SOLUTION: The most likely cause of this is human error. First check that the hutch is set and the safety shutters are in the open position (the 'shob' command). Next check that there is actually a beam from the beamline status display (the front end must be open/green and there must should be a significant ring current). Next ensure that the fast shutter is correctly set using the 'fshstate' command. The correct setting should be 'vct6' which can be set using the 'fshvct6' in SPEC. Finally, check to see if the detector is in the correct position (the 'wm trdet' command will give the detector position). Visually checking the detector position in the hutch is also possible. If the detector is in the wrong position, use the 'mv trdet xxx' command to move the detector to the 'xxx' position (which is whichever position it should be in).

If the above solutions do not remedy the problem, it is likely to be more serious and is unlikey to be fixed by user intervention. If this is the case inform the local contact (before 11pm).

There is low intensity

PROBLEM: There is some intensity reaching the sample/detector, but much less than anticipated. Thus a diffraction pattern or diffuse air scattering is still visible on the MARCCD display following a scan but the intensity is much less. Alternatively there may be a higher than background level readout from the ionisation chamber (counter 6), but a much lower value than was previously obtained.

SOLUTION: Firstly, a lower intensity may just be due to the decay in beam current although this should be gradual rather than instantaneous. Alternatively it may be that the scaling on the MARCCD display has been changed (so the image appears less intense). This can be adjusted by altering the upper and lower limit values on the MAR software. If these explanations are definately not the case, it is also possible that there has been a recent refill so the orbit of the ring has shifted. In this instance try running any alignment macros that the local contact has demonstrated.

Although this problem is far less serious than a complete loss of beam, there are so many causes that a full solution is not possible here. The exact solution will depend upon the setup being used and therefore a single solution or check-list cannot be given. Inform the local contact (before 11pm) for beam re-alignment.

The hutch interlock cannot be opened

PROBLEM: The hutch interlock cannot be opened using the open button on the hutch panel.

SOLUTION: This is probably because the safety shutters have not been closed, use the 'shcb' command to close the safety shutters and try again. Alternative, try depressing the button for longer. If the problem persists then inform the local contact (before 11pm) or floor operator (after 11pm) immediately.