Beamline X12-C Operation and Software Documentation

beamline X12-C operation and software documentation
px at nsls

Below will be found the documentation for operating beamline X12-C. The documentation for the running of data-reduction programs like denzo, Mosflm, or XDS are available elsewhere, though some Denzo hints can be found below.

If you encounter problems with the beamline, be sure to read Beamline Troubleshooting. You should take the time to familiarize yourself with the troubleshooting documentation before problems occur.


Your Visit to the X-Ray Crystallography Beamline.

Life at the NSLS:

There is a nice (and very complete) description of almost everything you might need to know about how to get along on Long Island and at BNL, provided courtesy of the BNL Public Affairs group.

All people who do work at BNL must have an "appointment" with the Lab. This will be a "guest" appointment and you will arrange it through the NSLS Users' Administration office. If you are not a US citizen, the process may be more complicated than if you are. The Biology Department User Coordinator's name, phone number, and E-mail address can be found on the front page of beamline documentation if you need forms or advice.

A second item of business is your Experimental Safety Form. You will already have given us information about the safety aspects of your experiment in the application form. The Biology Dept. User Coordinator will help you to submit a form; please be sure it's done in advance.

You can make your own arrangements in advance for housing with the BNL housing office (phone number: 631 344 2541). They're pretty convenient and inexpensive; they can be paid for with a credit card. You have a responsibility, once reservations have been made, to use them or cancel them. If you fail to cancel unused reservations, you'll have to pay.

Visitors are strongly encouraged to arrive for their visit to the NSLS early enough to spend some time checking crystals, getting training, assuring that all tools are in place, etc. An arrival at the lab of 3 PM or so will allow time to obtain help if necessary from the beamline technician before he goes home, and to get the NSLS training completed.

After your arrival at BNL you should go to the NSLS Users' Administration office to register. To find this office, enter the front door of the NSLS and climb the stairway behind the doorway to the left. After filling out various registration forms, you'll get a film badge and be instructed to watch a video tape about safety. We reiterate from above that we encourage you to come early. You could be taking data by 9 AM if you plan ahead.

At any hour of the day or night you can obtain your safety training through the personnel in the NSLS control room. They will give you a blue card-key that will admit you to the experimental floor. In the worst case after an early arrival one will have time and materials to make preparations or repairs that might have wasted hours of beamtime the next day. In the best case, one can go to the beach. Note: You do not need to register at the NSLS users office to begin your experiment if you have the safety training! The registration can be performed at leisure during the first day of running. Come straight to the beamline on the first morning.

Things that typically go wrong include having freezing loops mounted at the wrong height, having some crystals die so others need to be soaked, or not having the right goniometer-head key.

We remind users that the principal factor that contributes to lost beam time is crystals that degrade during transportation to BNL. Take very seriously the question of how best to transport the specimens. We recommend that you decide on a plan, prepare some crystals that way, drive them around in your car for a day, then see if they still diffract.

We have our own laboratory right at the beamline. You may manipulate specimens, think quietly, or do computing there (there's a PC). There is a coldroom nearby in the building for your use. You may have food at the beamline, but please take no food into the laboratory -- we're on shaky ground with the safety committee as it is. There is a key for the laboratory; ask where it is and keep it locked when you leave for an extended period.

The color TV monitor at the beamline tells you things about the operation the ring. Channel 9, the most useful channel, has a continuous message regarding the status of the x-ray ring. It's Red during injection, Green during ramping, and Blue during operation. You can read the present current and the 1/e lifetime, as well as some messages about operation. When calamities occur, up-to-date information about repairs to the synchrotron can often be found on channel 2. Channel 11 carries the weekly schedule for ring operations.

At the end of your visit, before you leave BNL, please submit an end-of-run form.

It's also crucial to the continued funding of the facility that you submit an abstract describing your result when we request them (typically once a year).

Go Back To Table Of Contents (TOP)


Summary of data-collection procedures

Written by Ruiming Xu and Margaret O'Gara

START UP THE SOFTWARE:

Operation and troubleshooting of the data collection software depends on the detector (B4,B1) being used.

B4 software summary

This is intended as a summary of the critical software aspects of the B4 detector. More details concerning the B4 are available.

The detector actually is made up of a 2x2 array of 1K CCD detectors. Each detector is controlled by a PC running Linux. Their nodenames are lsx12i, lsx12j, lsx12k, and lsx12l. On one of the beamline computer monitors should be four shell windows, each of which is a telnet session to one of the PCs. The username for the PCs is "det", we will give you the password. The only thing that should be run on the PCs is a progam called "server". After starting "server" and seeing some information scroll by, the program should wait with "Accepting service request". At that point the data collection program "CBASS" can be started.

B1 software summary

The single-celled CCD detector is controlled by a Windows PC that starts up the detector-control software at boot time. Users almost never need to touch this computer. More details concerning the B1 are available.

Starting CBASS

Before starting the data collection program, do a "df" to find a disk with sufficient free space and check the Disk Space by Beamline page for the beamline at which you are taking data. Data will be collected under the "x12cuser" subdirectory in each. Environment variables ($u2, $u7,...) point to the right places on the /usr* disks. Disks are always coming and going so you should basically just do a "df" and find a big drive with lots of space. All files should be stored in the user directory for the beamline you are working on.

After locating a disk with enough space, create a top level directory with a name that is desciptive of your institution, PI, or project. Move into that directory and type "cbass&".

After a few seconds the CBASS GUI will pop up with a message "downloading please stand by" in the scrolled white window at the bottom. After downloading is complete, a white shell window will pop up in the upper right hand corner of the screen if you are using one of the CCD detectors. This window contains the output of a "server" program running on an sgi that will service requests from CBASS for detector data. The "server" running on the sgi will then communicate with the "server"s running on the PCs to get the data. The text scrolling by in the white server window is informational and there is no user-interraction directly with that process.

ALIGN THE BEAM

SET THE WAVELENGTH:

CENTER A SPECIMEN:

OPERATION OF THE IMAGE-VIEW SOFTWARE

"smv" is a program that will automatically display the most recent image that has been recorded by the detector. We HIGHLY RECOMMEND that you keep it running at all times and that you LOOK AT YOUR DATA AS THEY ARE BEING MEASURED.

From a shell window on one of the workstations other than the one that you're using to run the data collection software, type "smv&" That will bring up a display program written by Marty Stanton at Brandeis University. Do the following to get it to display images automatically:

TAKE TEST EXPOSURES:

Choose a starting orientation for your crystal that you like and then press the Set Relative Zero to Current Position button in the lower left hand section of the CBASS GUI. This should not be done again with this crystal.

We suggest that you take test exposures in the Det level so that you can take advantage of the "fake dark" feature. When faking dark images, the software will use or take 2-second darks rather than requiring dark images of the same duration as the exposure. This can save time if you are varying the exposure time while screening crystals.

COLLECT DATA:

We recommend that you use our Strategy program to predict the most efficient data-collection range for a sweep of data or pre-orient the crystal. Suggestions for how to do this can be found below. However....

SOPHISTICATED MONOCHROMATOR CONTROL:

TO RUN "STRATEGY" OR TO PRE-ORIENT A CRYSTAL - DENZO Indexing with CBASS See also the detailed description of Orient below. The instructions below are intended to minimize errors. Experienced users will find that there are shortcuts. We recommend that you cut no corners until you have become familiar with the interplay among the images, crystal parameters, programs, and files involved in this process.

PAUSING DATA COLLECTION
You can pause data collection at anytime by pushing the Pause button in collect. After pushing the button, its label will change to Continue. Pushing Continue will resume data collection. You may use this feature to perform a Realign in OptiX during data collection or after a fill by performing the following steps:

ABORTING DATA COLLECTION
You can abort data collection by pushing the Abort button in collect.
You must not abort data collection if you are in the process of doing a wavelength scan!!

Go Back To Table Of Contects (TOP)


Odds and Ends of Operation

26 October 1994

Connecting Your Own PC to the Network:

Users may connect their own PC to the laboratory network through the DCHP server by following the instructions available on the NSLS web-site. The computer must be registered by filling out a web-based form within 30 minutes after the computer has connected to prevent the network port from being deactivated.

Specimen Cooler:

The FTS cool air system is no longer in regular use. If you have an application that requires this system in place of the Oxford cryogenic cooling system, please let us know in advance of your arrival either through your beamtime request or by direct e-mail to the technical staff.

A document that describes how to operate the controller for the Oxford Cryostream liquid-nitrogen-based cooler is kept in the file cabinet at the beamline, and some troubleshooting information is available on the web. One can find a brief description of how to get the unit going below.

Nitrogen Supply: There is a 50 liter dewar located inside the hutch from which the Oxford system uses liquid nitrogen. When full, this dewar will last for 48 hours of continuous operation. This "use dewar" is equipped with an auto-fill system (the "blue box," located in the coldstream equipment cart) supplied by a larger dewar outside of the hutch. The controller should always be left in the "AUTO" position unless a staff member tells you otherwise.

Setting the cryo-controller: To bring the Oxford Cryo-System on-line, first try to reach an operator or a member of the beamline staff (during the day). If you cannot locate a staff member, you may try to use the following instructions to bring the cooling system on line.
  1. Check that the power is OFF. The switch is located on the back of the controller.
  2. Then turn the power ON and wait for the system to complete its self checks. The screen should read "ready for use with cold head #xxx".
  3. Press the "program" button. The screen should change. Using the up or down arrow key, choose "RAMP" and press "enter".
  4. The cursor should be in the column "rate" now. Set the rate using the "fast arrow" keys to 300 k/hr and press "enter".
  5. The cursor should now be in the "temp" column. Using the arrow keys (or fast arrows) adjust the temperature to either 95Kelvin or 100Kelvin, and press the "enter" key.
  6. The cursor should be in the "function" column now (below "ramp"). Pressing "start" will make the controller bring the cooler to your set temperature. If you now press "screen" the display will change to normal operational display.

A crucial decision is how close to put the snout of the Cryostream to your specimen. We suggest that you bring the cooler down until you can see the shadow of its business end in the crystal-viewing microscope, then back it off until that shadow is just at the edge of the field of view.

We suggest that you put your crystal onto the goniometer head with the diffractometer set to the position equivalent in Eulerian angles to omega = 180 and chi = 45. In this way the crystal axis will be pointing 45 deg. downward, and it will be perpendicular to the stream of cold gas coming from the Oxford system. One can do this from the DET level or with the manual-control pod. The settings on the kappa goniometer that are equivalent to these Eulerian settings are omega = 160 and kappa = 60.

It may be necessary to withdraw the cooler from the position described above to mount your specimen -- bring it back no more than is absolutely necessary, then put it back as soon as the specimen is in place. When very high resolution data are being taken, there may be a shadow of the cooling snout on the detector. It is possible to remount the cooler slightly upstream, but that should be done by the beamline staff. Think about choosing a shorter wavelength.

Goniometer Heads and Height of Freezing Pins:

*A good, quick rule of thumb when using Hampton pins is that a pin broken off on the second "notch" will be about the right height to fit on our machines.

The FAST diffractometer has no z-axis adjustment possible. One must use a goniometer head with a built-in adjustment. The distance from the flat surface on the phi axis on which the goniometer head is mounted to the crystal position is 64mm. Other useful distances are shown in the diagram to the right and the one below.

We use Huber arc-less heads, which have a pretty long adjustment. The range of distances from the top of the platform on these heads to the specimen position is approximately 21 to 26.5 mm. For greatest success in centering crystals, shoot for a distance of 23 mm. from the platform to the crystal.

For the freezing of crystals, folks often use little steel cones or top-hats that fit onto a magnet affixed to the top of the goniometer head. We have a strong magnet approximately 3 mm thick with a pin just less than 3 mm diameter and approximately 1 mm long prodtruding from its top. Therefore, the height of a crystal-freezing loop should be between 18 and 23.5 mm above the bottom of such a steel cone, which should have a 3mm-diameter depression >1mm deep in its bottom.

Others like to use the proprietary Hampton Research Company system. We own one of their magnet assemblies, which can be placed on the top of a goniometer head. Important distances are shown in the diagram.

If you intend to bring your crystals pre-frozen, you should take very seriously the problem of making absolutely certain that you will be able to position your crystal perfectly on our diffractometer. Please bring a pin/top-hat of identical dimension to that used for your pre-frozen crystals for alignment of the goniometer head, etc.

X-Ray Shutters:

We like to keep the "Safety" shutter open continuously during data- collection runs. Access to the hutch, etc., can then be controlled with the Photon shutter. The reason to leave it open is to keep the optical components, especially the monochromator crystal, up at operating temperature. The safety shutter is under over-ride control by the NSLS control room. They will close it during the filling of the ring and reopen it when you see the "shuttrs enabled" message appear on the status screen. You will need likely to operate only the Photon-Shutter switch button.

There is a pair of shutters connected with the diffractometer. The diffractometer has a solenoid-driven shutter, the status of which is indicated by the Beam-On / Beam-Off lights right on the diffractometer. In addition there is a pneumatic shutter upstream of the diffractometer, mounted behind a large copper-plate x-ray shield. The internal timing of the operation of the machine is intended to have the very fast pneumatic shutter actually controlling the exposure. During normal operation, both are activated by Shutter-Open / Shutter-Close commands to CBASS.

Manual Control of Diffractometer:

The manual-control pod in the x-ray hutch allows one to drive the diffractometer by hand. It often pays to press the [clear] button when the device is picked up, just to empty its little brain. There is the possibility that the readout will be garbled although the commands work correctly. A beamline person can fix it for you.

There are two useful modes:

  1. To drive an axis to a preset position -- enter
    • [angle]
    • [move]
    • [axis -- omega, kappa, or phi]
  2. To drive an axis continuously, say to spin phi for crystal alignment --
    • [left arrow for negative motion] or [right arrow for positive]
    • [9] This is a speed selection.
    • [move]
    • [axis]
    • The axis will move until one hits [0] or [clear]. This provides a mechanism for easy control -- hit [0] to stop the axis; hit [9] to restart; hit [0] then the other arrow to change direction, then [9] to restart. Also, try the [pause] button.

Crystal-Visualization Microscope:

There is a video microscope mounted above the crystal position, viewing vertically downwards. There is a video monitor in easy viewing overhead at the diffractometer, and a slave outside. There are electronic cross-hairs to show the position of the rotation axis (single vertical wire) and the x-ray beam (double horizontal wires, between which the x-rays can be found). Illumination comes from a fibre-optic lamp suspended above the specimen. Note that it rarely helps to turn the lamp up above its lowest setting -- if it seems not bright enough, put it closer. Uninitiates are not encouraged to adjust the cross-wires, nor to try to focus the microscope. When in doubt, adjust the crystal to be stationary during rotation and ask your host for advice.

Baby Pictures:

One can use the video capabilities of the Indys at the beamline to take home a picture of your crystal:

  1. Login to "x12c-w"
  2. Type "capture" to start up the video-capturing tool
  3. Click the lower-left corner icon to choose still-camera icon
  4. Click "Actions.Settings" from the menu
    • Set the directory path and file name for the image.
    • If you turn on "Auto-Increment" the file-name will be changed after each image.
    • Check the "Control Panel" and select the "Default In" device to be the "Analog Source"
    • Set "Lock to VTR," Video Format to "Composite," and Input timing to NTSC (525)
  5. Back on the "Capture" window, click "Record"
  6. Type command "ipaste fimename" to see it.

Automatic Display of Images: Images are displayed using a program called "smv". More information can be found above.

Remarks on MAD-data collection:

A subtle point is how best to choose the "remote" wavelength. It should be sufficiently far from the inflection of absorption that the del-f' has increased as much as possible. A possible solution in the case of L-absorbers is to choose the point where the f' value has increased to its maximum between the L(III) and L(II) absorption edges. The figure at the right shows the example of the calculated spectrum for platinum atoms. However this position, which lies over a kilovolt above the L(III) edge, represents a 10% increase in the energy, or concomitant decrease in wavelength. The disadvantage of this in this context is that the curvature of the Ewald sphere will decrease by this same amount and the reflections that appear on the detector will differ for the same rotation range. For this reason, a reasonable compromise may be to increase the energy by 200 - 500 volts for the remote position.

For other edges and PostScript plots of other edges, see Ethan Merritt's data about anomalous scatterers: numbers and graphs .

Entering Collect Runs with a Sweeps File

When a data collection run starts, a "sweeps.dat" file is generated that records the information in the "runs" table of "collect". An identical file under a name that embeds the start and end positions is also generated. This file can be edited if needed and read back into CBASS from the "Files->Read" menu. This feature is useful for restarting/continuing a run that was interrupted for some reason. When editing the file there must be a directory entry but it is not used by collect. Collect only considers the file prefix and the current directory. The directory field is used when processing with "onebutton". You must be very careful not to alter the format of this file while editing.

Go Back To Table Of Contents (TOP)


DENZO Hints

Go Back To Table Of Contents (TOP)


Onebutton - BNL's Automatic Graphical DENZO Processing Tool


How to use it:

Typically (we strongly recommend this) one would create a sub-directory below the data directory for the data reduction.

There are two ways to use the program. Both require a "peaks.file" resulting from running HKL's xdisp program (formerly INST_something).

If you are currently taking data , then copy the files auto.dat, refine.dat, and sweeps.dat, that have been generated for you by CBASS, into your processing directory and use option 1.


Option 1) If you already have auto.dat and refine.dat in your processing directory:

type : onebutton (return key)

Running "onebutton" in this fashion requires "auto.dat" and "refine.dat" command files that are good enough to index the image that was used to generate "peaks.file".


Option 2) Uses a MAR image with correct file header info. (The ADSC MAR software usually takes care of this for you)

type : onebutton [mar image filename] (the one peaks.file was generated from)

In this case, the software should generate a pretty decent auto.dat program from this. However, if it is not good enough to index the image, you may want to get out of onebutton, make adjustments to "auto.dat" and then re-run onebutton as in option 1 (i.e. just type "onebutton").


In both cases, a file that CBASS generates to record data sweep info (sweeps.dat) is also used but will be generated if you do not have one. This file may be useful when users set up the data collection program to take multiple data sweeps, like when they're using the Friedel Flip to measure Bijvoet pairs, or doing a MAD experiment.


How "onebutton" works:

Given the above, "onebutton" will do the following:

1) Index and refine the image specified in auto.dat.

2) For each data sweep requested in the data collection program:

Generate command files, named "auto#n.dat", where "n" is the row number in our data collection window. Each will contain the refined parameters from the autoindexing run, such as distance and beam-center coordinates. At this point onebutton will prompt the user for more subjective parameters, including the spot diameter, mosaic spread, outer resolution limit, etc.

3) Run denzo with these command files while displaying a continuous graphical analysis of Denzo's output.

4) Generate textual log files for some of the graphical data.

Note:

You may pause and restart anytime before the last image is processed. To restart after all images have been processed, you should quit and re-execute the program.

Feel free to Download onebutton. The distribution contains everything you'll need including a few image files. Be sure to read the README file included in the distribution.

If you have comments, questions, complaints, etc., please contact:

John Skinner (skinner@bnl.gov)

Go Back To Table Of Contents (TOP)


Last modified 9 April 2002 by Protein Crystallography Web Master