Quick Fixes

Trouble Shooting on the Beamlines

 

 

	SHORTCUTS
Flat count rate on I0/I1	Sample Drop	Beamstop Misaligned
MXcube log-in failure	Fast Shutter Problems	Blank Images
Loss of Control from mxCuBE	Slit Motors Description	Cryostream is warming up
Magic Commands	ISPYB thumbnails not displayed

 

 

Can't log-in to MXcube

• because your proposal ID is not recognised

Your proposal number is new and MXcube doesn't recognise it and prevents you to log-in. This happens very rarely and is due to a failure in the updating of the ISPYB database. In order to validate your proposal you need to log-in first in ISPYB. Once this is done you should be able to log-in into MXcube.

If the problem persists, contact your local contact and he will log you in on the beamline generic account.

 

 

• you get the error message "DB server response in invalid"

 

 

The local ISPYB server is down. Ask the bliss on-call to restart it or if it is too late to get any help, when log-in in MXcube clear the proposal ID and use the beamline generic password as the password in MXcube. You should be able to carry your data collections but no parameter will be stored in the database.

 

Loss of control via mxCuBE

 

Occassionally you may observe that you can no longer control data collection and the beamline via mxCuBE. The usual problem here is a loss of communication with the spec session or the spec session has died. The most obvious diagnostic that this is the case is that the motor values as displayed in mxCuBE will have a red background instead of a green background. There are several things to try.

 

Find the spec session on the control pc desktop - is it still there?

 

If it is, then try to type "reconfig" in the spec session. If you regain control all the motor backgrounds in mxCuBE will turn green.

 

If not or the above reconfig hasn't worked then a clean start up. First use the "Stop spec" button on the desktop to clean up any old spec processes followed by clicking "Start spec" (or it's equivalent, eg "Start EH1"). If everything is fine the motor backgrounds should turn green.

 

If this doesn't work, quit mxCuBE (with the stop mxCuBE icon on the desktop), stop spec as described above, start spec and then start mxCuBE. Now everything should be working.

 

If none of the above works feel free to log out of the machine, log back in and restart all software.

 

 

Sample Drop

If your sample appears to droop just after being mounted by the sample changer or by hand :

• it can be due to ice on your cap melting once in contact with the warm magnet. Just wait for it to stabilise (few seconds) and then align the sample.
• it can be due to the pin not properly glued onto the basis.

 

 

 

Magic Commands

Here are some magic commmands in the experimental hutch spec sessions that may help solve a number of problems:

1. Hit return in the spec sessions.
2. Type reconfig. You can type this as many times as you wish. Twice is often a good number.Type resetup. You can type this as many times as you wish. Once is usually sufficient.
3. Type SCMoveToLoadingPosition in the experimental hutch spec session (usually called exp). This command is handy if the spec session is reporting that the sample changer is locked and you cannot load/unload a sample. Caution: afterwards you may lose control of the minidiff.
4. Type SCMinidiffGetControl in the experimental hutch spec session (usually called exp). This command is handy if the spec session is reporting that the minidiff is locked and you cannot centre your sample. Caution: afterwards you may lose control of the sample changer.
5. Type SCInUse in the experimental hutch spec session (usually called exp). This command is handy if the spec session is reporting that the minidiffractometer and sample changer are both locked. Caution: afterwards you may lose control of the sample changer.
6. Type phi_init in the experimental hutch spec session. If the phi axis is moving freely when you rotate it by hand, use this to find the zero position and relock to the axis - bearings. This doesn't work on 14-3 (microdiff instead of minidiff). On 14-3 close and restart the microdiff software running on the PC to the left of the beamline control pc.

 

 

 

Flat count rate

Symptoms of this are:

- flat lines at the heighest point during a quick realign due to saturation of the I0 counter.

This is easy to resolve by either repeating the quick realign, or if you are making a "ct" in spec, type "reconfig" or "wcadjall" in the experimental hutch spec session (called "Spec(exp)") with the safety shutter open for I0 and both safety shutter and fast shutter open for I1. This will readjust the gain on the counter to be something sensible.

 

Beamstop Misaligned

Symptoms of this are :

i) the diffraction image is blank when the diodes i0 and i1 are reading non-zero (checked by opening the safety shutter and fast shutter in the Hutch tab of the Data Collection GUI and doing a ct in Spec) OR
ii) rings are visible on the diffraction image.

1) Put the YAG screen in place with the crystal orthogonal to the beam, and focus to see the surface clearly.
2) Open Safety shutter and fast shutter.
3) Check there really is beam by putting the beamstop to the Out position.
4) Once you see that the beam is there, move the beamstop to the In position and open the beamsize up to 300um horizontally and vertically. This allows you to see the shadow of the collimator. This shadow should be round. If it is "lemon"-shaped then it may be necessary to adjust the vertical or horizontal tilt of the backstop, if you think the beam may be hitting it.
5) If you are happy with the shape of the collimator shadow, check that the collimator is properly centred around the beam position. If it isn't, click on the Beamstop align button on the Data Collection GUI and use the arrows to move the beamstop up/down/left/right until it is. Don't forget to save these positions before exiting.
6) If the collimator is "lemon"- shaped, you may adjust the horizontal tilt (vertical tilt more difficult, don't generally need to do this). To do this you need to adjust the screw keeping the backstop aligned on the support. Be aware that this screw is not very accurate and has to be moved with caution. Try moving 1/8th of a turn in one direction, and check on the YAG screen if the shadow looks rounder or not. If not, try 1/8th of a turn in the other direction. Then repeat 5).

 

Once the collimator is aligned, take a diffraction image while leaving the YAG screen on. Please use the same phi value in the data collection window as is displayed in the hutch tab. This should not show any rings anymore.

 

 

Fast Shutter Problems

Sometimes the fast shutter controllers get locked in a Protect mode. This stops them from opening and closing. Symptoms are either blank images (shutter stuck closed) or images with white diffraction spots (shutter is stuck open and so there are diffraction spots on the dark image).

To check this, first do a "ct" in Spec and note the values for i0 (before fast shutter) and i1 (after fast shutter). Open the safety and fast shutters in the Data Collection GUI and repeat a "ct". Both the i0 and i1 values should have increased notably. If only i0 changes and i1 is still the same it means the fast shutter is not moving. Close the safety shutter and go into the experimental hutch. Find the fast shutter controller box and using the switch on the front, turn it off and then on again. Make sure that the orange and red lights come on when you switch it back on. If they don't, switch off and on again until they do.

Repeat the above operation again. The shutter should now be working. Remember to retake a new dark image if you thought the shutter was stuck open.

 

 

Blank Images - General Diagnostics

If you get a blank image, the beam is not reaching the crystal. This can be due to MANY things and so it is necessary to perform some diagnostics to see where the problem is.

NOTE : Check the following obvious points first!!!!

i) Is the Front End open?
ii) Are the undulator gaps closed to sensible values?
iii) Are the optics and exp hutch actually interlocked?
iv) Are the gate valves open all the way along the beamline?

To check this, go to the Expert Front End desktop on the right hand screen of the control PC. There should be an application running there called xvacuum, showing the vacuum readings and gauges for the beamline. If it is not running, start it by clicking once on the "Vacuum Application" icon on the left hand screen of the control PC.
Check that all valves are green.
If they are white, they are closed.
To open the valve, click on it on the GUI and a small window will open.
Click Open and the Dismiss the window (DON'T click on Close as the closes the valve again!).
The valve should turn green. If it doesn't it means that the vacuum is not good enough to allow the beam onto the optical elements, or that there is a problem with the vacuum control. You should call the LC or the Experimental Hall Operator (25-25)

If the answer to those 3 questions is Yes, you need to perform the diagnostics by making use of the various fluorescent screens and counters installed on the beamlines.

First of all you need to check whether there is beam in the experimental hutch. To do this, use the counters i0 and i1 which are in the slitbox.

 

Counter i0 is situated AFTER the slits but BEFORE the fast shutter.
Counter i1 is situated AFTER the fast shutter but BEFORE the beamstop.

To check that the beam is coming into the slit box,

i) Make sure both the safety shutter and fast shutter are closed.
ii) Do a "ct" in Spec and note the values for i0 and i1. These are the values when there is no beam.
iii) Open both the safety shutter and the fast shutter using the Data Collection GUI, and redo a "ct" in Spec. Note the values of i0 and i1.

Diagnostics :

1) If the values of both i0 and i1 have notably increased (i.e. not just in noise) it means that there is beam through the slit box and past the fast shutter. A blank image is therefore likely to be caused by the beam not coming through the collimator of the beamstop. This implies that the beamstop has probably been knocked. Go into the exp. hutch and check that the beamstop is lying flat as it should. If it looks OK, it means that the beamstop needs realigning (see 2 above).

2) If the value of i0 has increased but the value of i1 remains the same, it means that there is beam after the slits but that it is not getting past the fast shutter. This means that the fast shutter is not opening. This may be because it has gone into Protect mode. Go into the exp hutch and check the fast shutter controller box. The Protect light should be off. If it is on, switch the controller off and then on again. When you switch it on the red and orange lights should come on - if they don't, switch off and on again until they light. Repeat the diagnostics above - if the fast shutter still doesn't open it means that there is a more serious problem with either the shutter itself or the controller - you will need qualified technical help for this!

3) If neither i0 nor i1 increase it means that there is no beam after the slits. This can be for a few reasons :

a) The slits are closed! Check the vertical and horizontal beamsize value on the DC GUI - they should both be non-zero.
b) The first and second slits have become misaligned. Call your LC!
c) All the attenuators are in! Check that the transmission value is not 0%. You can also go into the hutch and look at the attenuators themselves - when they are in the beam there is a red light. If all the attenuator lights are on then they are all in the beam, even if the software says they are not!
d) The safety shutter is not opening, check that SPEC (exp) is not reporting any errors and that the hutch is interlocked.

If you have checked all of the above and there is still no beam, it may mean there is a problem in the optics hutch. Various fluorescent screens exist to check this, e.g. flin1 and flin2 (use this command in the optics hutch spec sessions: mono_eh1 or mono_eh2 and the screens on shelves above the beamline control machines) on ID14-1 and ID14-2, mbv on ID23-1, ID29, etc.

 

 

For info - the slit motors for aligning the beam

The names are the Spec names!

 

And see the Slitbox page.

 

 

Cryostream is warming up

Check in the device control box if it is really warming up. If yes, then switch off the device (button at the back), switch it on again, push the "START" button from the device front panel after the initialization procedure, cool it down to 100 K.

 

My image thumbnails are not displayed in ISPYB

Check on this webpage the status of the JpegServer which is used. Each beamline has its own JpegServer and if it is in bad state you should contact your Local contact to get helped from BLISS to restart it.