X8-C

Hutch	Detector - ADSC Quantum 4 Diffractometer - Crystal Logic
Beamline
Workstation

X12-B

Hutch	Detector - ADSC Quantum 4 Diffractometer - Crystal Logic
Beamline
Workstation
Robot

X12-C

Hutch	Detector - ADSC Quantum 210 Diffractometer - Crystal Logic One should always check the vertical and horizontal cross-wires while starting a new group. If communication with the diffractometer's motors is lost for some reason, one has to turn the power off in the controller inside the hutch (the switch is right on top of the vacuum pu mp), and then turn it back again. The next step is to restart the Crystal Logic program on the PC. However, unlike, X25, there is no home switch in CBASS at present, hence one must HOME distance, vertical, and omega motors ONE AT A TIME. When the program is started, the z-translation for the goniometer is not enabled; it must be opened with an icon in that area. At present the best way to position the crystal in the beam is to rotate the crystal using the omega axis, shift from the camera plane to the XYZ plane and adjust x-translation when omega is 0 deg or 180 deg, and y-translation when omega is at 90 or 270. If you are moving the axes using the + and - icons, the axes need not be exactly positioned at these angles, but may be within 5 degrees of the numbers given above.
Beamline
Workstation

X25

Hutch	Detector - ADSC Quantum 315 Diffractometer - Crystal Logic
Beamline
Workstation

X26-C

Hutch	Detector - ADSC Quantum 4 Diffractometer - Nonius Collection is stuck on 'Readout' or 'Dark' Lost communications with Detector Computer. Abort collection. Close Cbass. Go to Detector Computer rack and restart 'Quantum Console' software. Restart Cbass. Goniometer is stuck at limits and not moving Last resort to move a stuck motor using the nonius control pad Press 'Left' arrow or 'Right' arrow (sets direction). Press '9' (sets speed). Press 'Move'. Press axis to rotate about (kappa, omega, phi). Keep your finger on 'Clear' button to stop motor motion. Moving omega, kappa or phi manually If the goniometer runs into a soft limit or bangs into the collimator, you may need to manually move the motors back to their zero positions. For omega, you need to remove the case behind the arm holding the detector. You will see two little gold gears, turning the outer one will move omega. For kappa, you need to remove the see-thru black/grey case covering the motor. For the three cases on the goniometer that fit this description, open the two smaller ones. You wil find that there is a dial which you can turn by hand on of them; this will move kappa
Beamline
Workstation

X29

Hutch	Detector - ADSC Quantum 315 Diffractometer - Crystal Logic Other X29 things Channel 15 posts gap seperation ssh x29-h, cd /img09/data1/pxuser, cbass & min D=89, max D=710 Home Omega before interlocking hutch Distance crystal needs to be from end of silver base should be adjusted to 40 mm
Beamline
Workstation	Rules of Thumb Cbass should be the last program you start at the beamline
Robot

Cryostreams

Cryostreams at: X8c, X12b, X25, X26c Cryojet at: X12c New Cryostream at: X29 Spare for X12c in X12c lab Spare for other beamlines in X26c lab

Data Backup

There are 4 options to back up your data from our data disks: DVD Backup Instructions are available here Backup to Laptop Instructions are available here FTP Instructions are available here Firewire To copy data from our disks to your data storage device, type: rsync -av --modify-window=1 [source directory] [destination directory] This will let you copy files and gives you the added ability to update new files by issuing the command again. rsync copies any new files that came into the source directory and does not overwrite anything already backed up. For information on formating and partitioning external harddrives, go to Reformat Instructions. "Recent addition (8/28/06) from (Matt) for Firewire procedures" Check and make sure the system doesn't already have any firewire/usb disks physically attached; if it does your best off using another system rather than playing with multiple disks on a single system, one of which is likely already in use. If the users are really adamant about wanting their disk on that machine, and the other disk is theirs too, it should work fine... but gets trickier as soon as there's any problem and you have to start being careful about not impacting the in use disk. In which case, you **should not** just verbatim follow the directions below as that will likely screw up an in use disk. -physically look at the system -don't forget to look at the firewire/usb ports on the front AND the back of the system. Check and make sure the system doesn't still *think* it has a firewire/usb disk mounted. -'df -l' -Note: check the output...is it /dev/hda(something) or /dev/sda(something) umount any ghost disk...you're SURE it's a ghost disk right? 'fuser -vumk /(whatever)/1394_data(XXX)' 'umount -f /(whatever)/1394_data(XXX)' If it doesn't want to umount, proceed to step (4). Check and make sure the system doesn't still think it has a firewire/usb disk attached. -'cat /proc/scsi/scsi' -Note: the gx280's, and newer systems have serial ata system disks, which show up as scsi, and which you would have seen in the 'df -l' above, and so you should see a single entry like: Host: scsi0 Channel: 00 Id: 00 Lun: 00 Vendor: ATA Model: ST340014AS Rev: 8.12 Type: Direct-Access ANSI SCSI revision: 05 Clean up any old garbage ghost or unmounted devices a) first try 'rescan-scsi-bus.sh -r' b) if that doesn't work, try unloading and reloading the appropriate drivers (usb_storage for usb, sbp2 for firewire) 'modprobe -r sbp2' (-r to remove) 'modprobe -a sbp2' (-a to add) c) if that doesn't work, you might actually need a reboot. Once the system is in a clean state, power your new disk up, and then attach it to the system; power up first, then attach. See if the system sees the disk... cat /proc/scsi/scsi If it sees it, check the partitioning (keep in mind that as mentioned above some of our newer systems have an internal /dev/sda) cat /proc/partitions -(if all is going well so far and you haven't already needed to be root, you probably can skip checking w/ fdisk and parted, if you have a problem, the first thing I would do would be to check fdisk and parted) fdisk -l /dev/sd(whatever) parted /dev/sd(whatever) print Those should all agree otherwise something is wrong and go to step(9). Hopefully it's just got 1 fat32 partition...just mount it and your done. If it's got wackier partitioning, you'll need to figure out the correct fstab entry to mount depending on the partition# and type (fat32 or linux) you want to mount, you might have to make new fstab entries for unusual partitioning schemes. If it's linux format you'll also need to ensure that a directory on the disk is writeable by pxuser. I usually just make a new directory, 'mkdir /(whatever)/1394_data(xxx)/nsls-x29', or whatever and 'chmod a+rwx /(whatever)/1394_data(xxx)/nsls-x29'. If it doesn't try 4{a,b,c}) in that order until it appears in /proc/scsi/scsi, then go back to step (7). Call Matt If you have to (re)format at disk, we already have nice instructions for that on the website. Reformat Instructions. When setting up a firewire, follow this procedure: It would be useful to double check what a machine thinks it has mounted to it.Do a 'cat /proc/scsi/scsi' to check if there are any scsi devices attached. To release a scsi port: As Root: echo "scsi remove-single-device 2 0 0 0"> /proc/scsi/scsi where the number '2' designates which scsi port is occupied (get this number from 'cat /proc/scsi/scsi'). After verifying there is no scsi ports occupied, run cat /proc/scsi/scsi to make sure there is no attached devices. After verifying there is no attached devices, power on firewire and plug into USB2 or firewire port. (If you are mounting a firewire to X25-a, you need to plug your USB2 or firewire cable in Before you power up the firewire disk). At this point, just make sure that the firewire device has been recognized by the computer. Do a cat /proc/scsi/scsi to see if your device shows up. If it doesn't, then, as root, do a rescan-scsi-bus.sh. This should bring it up. After this, just type 'mount /computer_name/1394_data1', where 'computer_name' is the name of the machine you're mounting the disk to. Some quick remiders for firewire and such. Output from the following commands is useful. df -l dmesg cat /proc/scsi/scsi cat /proc/partitions grep kernel /var/log/messages (root only) "I/O error" or "File system read only" Firewire was unplugged before unmounted run 'fdisk -l /dev/sda' to see disk and partitions umount disk as root run 'dosfsck -av /dev/sd??' 'rescan-scsi-bus.sh -r' removes disconnected disks that may still be in 'df' or '/proc/partitions' As a last resort, you can reset the firewire driver by: /modprobe -r sbp2; /modprobe -a sbp2 To look at a list of the modules: /sbin/lsmod "Permission denied" 'chmod 777' changes ownership to read,write,execute for you "Device is Busy" run 'pwd' are you in the device? run 'sbin/fuser -vum /x__-_/1394_data1' to see what is using it, and kill it or -vumk option to kill automatically If the previous method does not work, due to that speedy gam_server, try running '/sbin/fuser -vumk /x__-_/1394_data1 ; umount /x__-_/1394_data1' this should kill it automatically and unmount the device before the daemon has a chance to fire it back up. "No Such Device" run 'cat /proc/scsi/scsi' is the device there? run 'rescan-scsi-bus.sh' "Recent addition from (Matt) for Firewire procedures"

Networking

Proxy Configuration To reconfigure proxies for Mozilla Go to the 'Edit' Menu and choose 'Preferences'. Find the 'Advanced' category and choose 'Proxies'. At this point, you can manually or automatically configure the proxy. For manual configuration, enter under HTTP Proxy 192.168.1.3 and Port 3128 For automatic config, type the address http://security.bnl.gov.proxy/cfg.pac Check Printer Status To check on printer status, in web browser, type http://w04.nsls.bnl.gov:631 This brings up the print server for all the linux machines.

Detectors

Shutdown NOTE: This list is intended as a quick reference/reminder and assumes a level of familiarity much higher than zero! Q315/Q315r (please refer to ADSC manual if possible!) On all 9 PC's, switch from 'Remote Operate Detector' to 'Operate Detector'. Ramp detector CCD's to +10C via 'Quantum Console' on the zeroth machine. After all CCD's have stabilized, turn off cooling power on the controllers (top to bottom). Turn off detector power on the controllers (top to bottom). Shut down detector PC's (9). **If power outage is expected, or if detector is to be moved; Close detector vacuum valve. Turn off main power on the Q315 rack. Power off the international power supply. Q4/Q4R (please refer to ADSC manual if possible!) Warm the detector by turning the knob in 10 degree increments until zero is reached. Wait 10 minutes AFTER the detector controller stabilizes (light turns green) before proceeding to the next increment. NOTE: the degrees indicated on the knob are NEGATIVE (i.e. '20' means '-20C'). Shut down controller cooling (top to bottom). Shut down controller power (top to bottom). Shut down computers. **If power outage is expected, or if detector is to be moved; Close detector vacuum valve. Turn off main power on the Q4 rack. Q210 (please refer to ADSC manual if possible!) On all 4 PC's, switch from 'Remote Operate Detector' to 'Operate Detector'. Ramp detector CCD's to +10C via 'Quantum Console' on the zeroth machine. After all CCD's have stabilized, turn off cooling power on the controllers. Turn off detector power on the controllers. Shut down detector PC's (4). **If power outage is expected, or if detector is to be moved; Close detector vacuum valve. Turn off main power on the Q210 rack.

Startup NOTE: This list is intended as a quick reference/reminder and assumes a level of familiarity much higher than zero! Q315/Q315r (please refer to ADSC manual if possible!) Power on international power supply. Re-program international power supply (follow detailed instructions on quick reference card... two person rule!) Power on main power on the Q315(r) rack. Check that vacuum pump is running. When line vacuum is better than detector vacuum, open detector valve... two person rule! Press yellow startup button. Chiller should turn on. Check that chiller is on. Wait until it reaches operating temperature... two person rule! Power on detector PC's. DANGER!!! Detector controllers must NEVER be turned on unless the chiller is at operating temperature!!! Turn on detector power on the controllers (bottom to top). Turn on cooling power on the controllers (bottom to top). Start 'Operate Detector' on all 9 PC's. RAMP detector (via 'Quantum Console') to 0C. Wait for all CCD's to stabilize. RAMP detector to -45C. Wait for all CCD's to stabilize. Close 'Quantum Console', and open 'Remote Operate Detector' on all 9 PC's. Q4/Q4R (please refer to ADSC manual if possible!) Turn on Main Power. Check that vacuum pump is running (switch is at the ON position). When line vacuum is better than detector vacuum, open detector valve... two person rule! Confirm that the rocker switch on the chiller is ON. Press yellow startup button. Chiller should turn on. Check that chiller is on. Wait until it reaches operating temperature... two person rule! Power on detector PC's. DANGER!!! Detector controllers must NEVER be turned on unless the chiller is at operating temperature!!! Power on detector controllers (bottom to top). Open 'Remote Operate Detector' on all 4 PC's. Q210 (please refer to ADSC manual if possible!) Turn on Main Power Check that vacuum pump is running. When line vacuum is better than detector vacuum, open detector valve... two person rule! Press yellow startup button. Chiller should turn on. Check that chiller is on. Wait until it reaches operating temperature... two person rule! Power on detector PC's. DANGER!!! Detector controllers must NEVER be turned on unless the chiller is at operating temperature!!! Turn on detector power on the controllers (bottom to top). Turn on cooling power on the controllers (bottom to top). Open 'Operate Detector' on all 4 PC's. RAMP detector (via 'Quantum Console') to 0C. Wait for all CCD's to stabilize. Ramp detector to -43C. Wait for all CCD's to stabilize. Close 'Quantum Console' and open 'Remote Operate Detector' on all 4 PC's.

Reset serial number If the detector PC's are powered off before the detector controllers are powered off, the serial number to the controller will probably be lost. First confirm this by running Dbglx32.exe file-> statusframe&voltages. Then fix, by proceedure below. Turn the controller power ON. Run the Dbglx32.exe program (there should be a desktop icon). Go to file -> fixXXXX (appropriate fix script). Run script. Power controller OFF, wait fifteen seconds, power controller ON. On Dbglx32.exe, go to file -> statusframe&voltages, and confirm that the serial number has been fixed.

Swapping in a spare PC (Q315/Q315r/Q210) Follow this procedure if you need to replace a failed detector PC with a spare. Warm up the detector (follow procedure). Power OFF the detector controllers. Shut down the computers. Swap all PC cables (network, fiberoptics, etc). Power on computers. Re-name spare to whatever the failed machine was (i.e. 906-3) Re-assign the IP address of the spare to whtever the failed machine was. If the failed machine was the zeroth computer, you must also make sure the appropriate files are on the -d drive of the spare (d:/pv_sw). If zeroth, make sure the -d drive is properly 'shared as' the machine name. Reset the spare's serial number.

Swapping in a spare PC (Q4/Q4R) Follow this procedure if you need to replace a failed detector PC with a spare. Warm detector (follow procedure). Power off controllers Swap cables on failed computer with spare. Change spare computer name to match failed machine. Change IP address of spare computer to match failed machine. Follow power on procedure.

Swapping in a spare controller (Q4/Q4R) Follow this procedure if you need to replace a failed detector controller with a spare. Shut down detector and controllers (follow proceedure outlined above). Disconnect 'detector to controller' cables at DETECTOR END FIRST! Cap off all connectors on detector. Disconnect 'detector to controller' cable at controller end. Disconnect all other controller cables. Swap in new controller. Reconnect all cables on controller end, including 'detector to controller' cable. Remove caps, and reconnect all cables at detector. Power on detector/controllers (follow proceedure).

Computer upgrade of Q315 (05/08) In May 08, we upgraded the detector computers for the Q315 at x25. The immediate result was a reduction in HW_bin mode, of the readout time from 2.X sec to 1.4 sec. Added a second CPU to the dual-CPU capable motherboard. Doubled the amount of stock memory (from 1GB to 2GB). Upgraded the OS from WinNT to WinXP in order to make use of the dual CPU. Copied the previous file structure, and installed the latest version of the ADXV software.

Hardware, Computer upgrade of Q315->Q315r (05/08) In May 08, we sent the Q315 at x29 to ADSC for the -r upgrade. Some notes follow. File structure is much the same. OS is now WinXP. HW_bin readout time is now ~1.1sec. Signal to noise ratio is much improved. Dynamic range is lowered (a 65k saturation now corresponds to a lower number of x-ray counts). The result of 3 and 4 may be an increased incidence of overloaded spots at low resolution. If this is the case, advise users to break their collection up into two sweeps, one high resolution, and a second with an attenuated beam to catch low-res spots. The two data sets are to be merged.

Misc. Stuff

Unix/Linux Commands: To remount all disks, type (as root), mount -a -t nfs To use scp: scp -r /source_directory /destination_directory This lets you file transfer pretty easily between two machines. Ion Chamber Operation (_ )=recommended values Plug in Change "Display" to preset Use "Select" & "Advance" to change count time multiple (10*10^1) Change "Display" to A Select time base (0.01) Press "Count" to see time is correct Adjust dwell knob for pause between counts Change "Display" to B for counts Change IC Plus voltage & gain (V=400,G=4)