Leica S430 Scanning Electron Microscope
- 1 Configuration
- 2 Authorization
- 3 Suppliers
- 4 Supplies
- 5 Documentation
- 6 Troubleshooting
- 7 Maintenance Log
- Secondary electron detector (functional)
- 4 quadrant silicon backscatter detector (non-functional, troubleshooting in progress)
- Oxford Isis EDX detector (working, requires liquid nitrogen to operate)
- High Resolution Recording Unit (HRRU) (not currently connected, Polaroid film no longer produced)
- Ethernet. Scripts set up to FTP image files for each user to ftp://semsrv.pumpingstationone.org which is accessible only inside PS:One's network.
This is probably the most complex piece of equipment we own. As such, authorization will be broken into different levels which qualify a user to perform only certain specific tasks. Unlike other PS:One equipment where anyone authorized can, in turn, authorize anyone else, at present the only person who may authorize others is Ryan Pierce. We anticipate we will have additional trainers in the future.
Scanning Electron Microscope User
This level enables the user to observe samples already prepared and placed in the chamber. Users at this certification level may:
- Observe samples already left in the chamber.
- Use the Secondary Electron detector.
- Observe their own samples if prepared and placed in the chamber by someone with Sample Preparer certification.
- Adjust standard software settings.
- Align the aperture, and change between apertures.
- Vent the chamber and restart the vacuum pumps. (only if necessary)
Users possessing this authorization may not:
- Use the Backscatter or EDX detectors.
- Open the chamber and manipulate samples.
- Adjust gun alignment.
- Open the gun assembly, change filaments, anodes, etc.
- Replace apertures.
|Qualified Member||Trained By|
|Ryan Pierce||Ryan Pierce|
|Steve Finkelman||Ryan Pierce|
|Tucker Tomlinson||Ryan Pierce|
|David Fell||Ryan Pierce|
|Tony Coleman||Ryan Pierce|
|Lisa Coleman||Ryan Pierce|
|Patrick Schless||Ryan Pierce|
|Everett Wilson||Ryan Pierce|
|Elizabeth Koprucki||Ryan Pierce|
|Andreas Dahl-Hansen||Ryan Pierce|
|Justin Conroy||Ryan Pierce|
|Dmitry Baycharov||Ryan Pierce|
|Jerry Lebedowych||Ryan Pierce|
|W J Perry||Ryan Pierce|
|Jay Hopkins||Ryan Pierce|
|Aaron Carlock||Ryan Pierce|
|Andrew Kos||Ryan Pierce|
|Todd Allen||Ryan Pierce|
|John Rudnick||Ryan Pierce|
|Cameron Rudnick||Ryan Pierce|
|Mike Warot||Ryan Pierce|
|Dave Schact||Ryan Pierce|
|A. Catherine Noon||Ryan Pierce|
|Arturo Duarte||Ryan Pierce|
|Melody Snyder||Ryan Pierce|
|Scott Anamizu||Ryan Pierce|
Backscatter Detector User
Users possessing this authorization may:
- Move the four quadrant backscatter detector in and out of the beam path.
- Image samples with the 4QBSD.
As the 4QBSD is currently undergoing maintenance, this authorization is currently on hold. It will be rolled into the Microscope User authorization and made part of that curriculum at a future date.
EDX Detector User
Users possessing this authorization may:
- Operate the EDX detector.
- Fill the EDX detector with liquid nitrogen.
PS:One only recently acquired equipment for transporting, storing, and dispensing liquid nitrogen. An EDX authorization class is forthcoming.
Users possessing this authorization may:
- Mount samples on stubs.
- Use the Technics Hummer V Sputter Coater to sputter coat sample stubs.
- Vent the SEM chamber.
- Load and unload samples from the chamber.
- Pump down the SEM chamber.
This authorization is currently on hold, but we anticipate to offer it shortly.
Maintainers perform all maintenance on the microscope, including:
- Changing filaments
- Aligning the gun
- Replacing the anode
- Cleaning the gun assembly
- Replacing apertures
- Adjusting software configuration
- Anything involving disassembly of the microscope
|Qualified Member||Trained By|
|Ryan Pierce||Ryan Pierce|
- Front panel light bulbs: These use #85 28V miniature bulbs. The scope came with many of these.
- Edwards EMF20 oil mist and odor filters:  has the best price I could find. These are two separate elements, and one may need to be replaced while the other may be fine.
- Leica S430 Scanning Electron Microscope Operations Checklist
- SEM User Certification (Available on internal network only!)
- Leica S430 Scanning Electron Microscope User Certification Checklist
- S430 Manual: Chapter 1
- S430 Manual: Chapter 2
- S430 Manual: Chapter 3
- S430 Manual: Chapter 4
- S430 Manual: Chapter 5
- S430 Manual: Chapter 6
- S430 Manual: Appendices
SEMSRV is the FTP server. The SEM and the EDX are Windows 3.x era machines, and are currently configured via BOOTP. The current Mikrotic router does not send DNS server information with its BOOTP responses.
|SEMSRV (virtual Linux FTP server)||1E:0B:6B:E6:2D:9C||10.100.0.108|
Final vacuum doesn't want to drop beyond 1.8e-5 torr. However, after leaving the scope on, the vacuum increases to around 1.3e-4 torr. Venting and pumping the chamber always fixes the problem. Also, when pumping down, the pressure bounces around a bit initially, and then suddenly drops from 1.0e-4 to 6e-5 torr within a second, which is likely a physical impossibility. Possibilities:
- The Penning vacuum gauge can become plated with ions. This is a very common failure mode and can cause false readings. Lightly sanding the anode and cathode, and washing everything in acetone, is the manufacturer's recommended cleaning procedure. This probably hasn't been done in a very long time and is a good first step before further troubleshooting. For all we know, we don't have a problem. Lending support to this is an absence of image degradation between the 1.3e-4 "bad" state and the 3e-5 "good" state after venting and pumping.
- The vent solenoid has been known to leak on other SEM users' scopes.
- Chamber and gun gaskets should be greased with a suitable high vacuum grease, e.g. Apiezon M.
- Rotation speed on the turbomolecular pump can be checked, which correlates to chamber pressure.
Cleaning the Penning gauge now results in a significantly more stable vacuum reading, with no wild fluctuations that defy the laws of physics. It had significant accumulated plating on the anode and cathodes, and loose black shiny grit in the gauge, which could cause erroneous readings. However, the gauge doesn't indicate below 1.09e-4 torr. Thoughts:
- Contamination in the Penning gauge often results in less current flow, which registers as a lower vacuum reading than is actually present.
- It is possible that we really have been running at 1e-4 torr all this time without realizing it. This is certainly not ideal, but a beam can pass through the column and image at this level.
- While the chamber air admit solenoid may be a possible culprit, this seems unlikely. Air leaks usually are larger than this. Still, plugging the solenoid is an easy way to check.
- The turbomolecular pump is now a likely suspect. 1e-4 torr is only marginally better than a roughing pump can achieve. Turbopumps should drop the pressure into the e-6 range. It is possible the pump is not running at full speed. This should be investigated via the pump controller.
- Applying light finger pressure to the air admit solenoid causes the chamber pressure to drop from 9.3e-5 torr to as much as 9.0e-5 torr; releasing it causes it to return to the starting value. This is repeatable. It suggests a possible leak related to the air admit solenoid.
After removing the Penning gauge, cleaning the O ring of residual black sand grit, and reassembling, pressures in the 9e-5 torr range can be achieved. Attention turned to the Edwards EXT120E turbomolecular pump controller. This has an analog output from 0-5 V proportional to pump speed on pins 16 and 17. (Polarity appears backwards from what the manual states.) I measured 4.955V, corresponding to the pump running at 99.1% speed. Thoughts:
- It is therefore very unlikely that we have bad bearings causing the pump not to run at full speed.
- Tightening the Penning gauge did produce a change in vacuum of around 5%. However merely rotating the magnets, which are in an external sleeve, changes the vacuum too. It looks unlikely to be a leak in the gauge. Rather, the gauge seems to be very dependent on physical orientation, even though, if left alone, the results are now very stable.
- Pushing the pump around when running did not produce a change in vacuum.
- Pushing around the solenoid valve did produce a change in vacuum.
- One of the bolts holding the solenoid on was loose. Tightening it didn't seem to be the fix.
- It is likely the solenoid valve leaking. Replacing the solenoid valve with a rubber stopper is the next step to troubleshoot the problem.
Solenoid valve was removed and replaced with a rubber stopper. This clearly fixed the problem, as low 1.0e-5 torr vacuum could be achieved. Attempts to repair the old solenoid were unsuccessful. Solenoid was replaced with a new old stock Edwards C41723000 24V NO solenoid. This required hacking together an NW10 adapter to connect to the air hose leading to the desiccant dryer / filter, and obtaining an NW10 centering ring and clamp.
This adequately solves our vacuum issues. 1e-5 torr is acceptable. In the future, we may attempt treating rubber O rings with vacuum grease in an attempt to improve the final vacuum into the e-6 torr range.
Image Processor board loses communications
We have observed after moving the scope to its new location that the vacuum spontaneously shuts off. Error messages pop up on the console several seconds later concerning timeouts to the Stage and Vacuum board.
Communications between boards happens via RS232 over fiber optic cables. The Opto 10 and 11 connectors handle communication to the Stage and Vacuum board, and unfortunately it appears tension was applied to the fiber, causing damage to the connector on the image processor board. The cable isn't locking in place properly.
On a single occasion, communication to the EHT board was lost. This corresponds to Opto 8 and 9.
A dual cable plugs into a single port, Opto 1, for the HRRU pixel data. The fiber plug connector on the cable end has snapped. This is not critical as we are not using the HRRU.
Repair is possible. The connector in question is a "hp 9534 R-2521" with a T-1521 next to it. These appear commonly available, but soldering to the image processor board should be considered a last resort.
We have removed the external fiber cables tangled with the others. This has largely reduced the incidents of vacuum crashes due to communications failures, but not entirely eliminated them. One failure has happened since then.
|10/30/2014||Ryan Pierce||Shut down SEM for repair of the rotary vane pump. Vented chamber. Removed all samples. Disconnected rotary vane pump. Disconnected electrical power. The rotary vane pump is exhibiting signs of overheating, scorching the oil, making very loud, regular clacking noises in time with the pump stroke, emitting a scorched oil odor, possibly allowing oil vapors to escape back up the line when the pump is turned off, and may have a small oil leak.|
|9/20/2014||Ryan Pierce||Finished vacuum system repair. Had previously replaced the air admit solenoid with an Edwards C41723000 24V NO solenoid. This produced a vacuum around 1.03e-5 torr. Added custom made adapter from NW10 flange to the fitting on the old solenoid that attached to the air dryer. This adapter was made by cutting apart the old solenoid, TIG welding a pipe compression fitting, which mechanically held the hose fitting. This was not airtight, so was sealed with epoxy. Chamber now vents properly and holds a vacuum. Re-assembled desk. Dried silica gel in an electric oven at slightly under 200F for several hours in a baking dish, stirring periodically, until bright blue color returned, and re-assembled air dryer assembly. Noted oil in rotary vacuum pump was running low, topped it off.|
|6/15/2014||Ryan Pierce||Replaced filament. Prior filament had 52 hours. Suspect user error setting current too high, however the filament was operating for most of its life at a very poor vacuum. Also changed out Weinhelt aperture with used spare. Old one needs to be cleaned and polished, not exactly sure how. Need to contact Susan (former operator.) At 10 kV, 1000x, 90 pA I Probe, first peak at 2.61A, second peak at 2.86A. Also corrected connection of mist and odor filter to rotary vacuum pump, which was not properly seated.|
|5/?/2014||Ryan Pierce||Removed chamber air admit solenoid. Replaced with rubber stopper. Vacuum ultimately gets to 9.6e-6 torr.|
|5/18/2014||Ryan Pierce||Noted turbomolecular pump analog speed voltage at 4.955 V, corresponding to the pump running at 99.1% speed.|
|5/1/2014||Ryan Pierce||Disassembled Penning gauge, straightened anode wire better, noted some of the particulates on the O ring, cleaned it better with a kimwipe, and reassembled. System pressure is now 9.3e-5 torr, which is sufficient to run the beam.|
|4/29/2014||Ryan Pierce||Disassembled Penning gauge, straightened anode wire better, cleaned O ring with kimwipe, and reassembled. System pressure is lower, measured at 1.09e-4 torr on 5/1/2014.|
|4/26-27/2014||Ryan Pierce and Elizabeth Koprucki||Disassembled Penning gauge. It was filthy with plated material and loose shiny black sand-like material. Sanded anode and cathode cups to remove the bulk of plated material. Removed ion trap and sanded interior metal surfaces. (The plastic surfaces are filthy, but I don't know how to clean this.) Cleaned all metal parts in 3 progressive changes of acetone in an ultrasonic bath. Anode wire was bent in re-insertion. Following re-assembly, pressure was 1.5e-4 torr.|
|4/7/2014||Ryan Pierce||Discovered that tension on the HRRU fiber cables snapped the HRRU fiber connector of one cable and damaged the locking mechanism on OPTO 11, which likely caused the loss of vacuum and loss of communication to the vacuum board. Removed the HRRU cables so they are no longer a hazard. Re-seated all fiber connections to the image processor board.|
|3/6/2014||Ryan Pierce||Replaced vacuum mist and odor filter, D ring, and drain plug gasket. Replaced front panel standby light bulb.|
|3/2/2014||Ryan Pierce||Changed CUTCP CONFIG.TEL file to use hard-coded DNS because the new router does not send DNS info with BOOTP responses.|
|1/7/2014||Ryan Pierce||Noticed HDD controller error. This has happened in the past. Physically moving the IDE cable fixes it. Moved the cable this time, and the problem went away. Suspect bad cable. So I replaced the IDE cable.|
|12/24/2013||Ryan Pierce||Changed all user scripts to point FTP to semsrv.pumpingstationone.org so that the old server, plank, can be retired. NOTE: While CUTCP does get DNS from BOOTP, the NCSA Telnet I'm using for FTP does not! I had to hard-code the IP address of the DNS (used 10.100.0.1) in the CONFIG.TEL file which is a nasty but necessary hack.|
|11/9/2013||Ryan Pierce||Changed rotary vacuum pump with oil that came with the SEM. Noted that the oil mist filter was dry, but not working well as a lot of odor is passing through. Needs new filter elements.|
|10/17/2013||Ryan Pierce||SEM claimed filament had blown when this was not true. Opened gun, put same filament back. Re-aligned. This was after 6.9 hours filament time.|
|9/26/2013||Ryan Pierce||Filament blew after 49 hours run time. Replaced.|
|9/12/2013||Ryan Pierce||Discovered that final apertures were set at 20, 50, and 100 micron. However, the spare parts that came with the SEM only had 20, 30, and 50 micron apertures. Reconfigured dialogue box to use these values and for the #2 aperture, the image become much clearer. So this appears to be a config error I corrected.|
|9/2/2013||Ryan Pierce||Installed NCSA Telnet (in addition to CUTCP package I already installed) because NCSA FTP supports scripted input. Set up user scripts and icons to push files for each user to the FTP server, and to delete files.|
|8/31/2013||Ryan Pierce||Installed network software in C:\NETWORK. Set up AUTOEXEC.BAT to launch DOS networking from startup. Used EZSETUP.EXE to configure Ethernet card for 0x280, IRQ 3, 0xCC00 as recommended in Leica manual. (Note: Memory address conflicts on factory setting. Use soft config jumper.) Set up DOS packet driver. Set up for BOOTP. Networking under DOS works, and the DHCP server will respond with the appropriate static IP. Installed WINPKT.COM which is necessary to allow apps under Windows to use the packet driver. Tested apps in a DOS box; work successfully. Moved unnecessary startup items to a new group so they do not launch on startup and consume RAM.|
|8/30/2013||Ryan Pierce||Changed from backup hard drive to system hard drive. (Backup that Jay made works. Note that he must have taken the backup when the SEM was in a weird state; you will need to load one of the configs in the default user directory just to see beam and an image. The LUT may have the colors inverted too. Reminder: have Jay make another backup once we're good to go with networking.) Removed laser printer card. Installed SMC Elite 16 Ethernet card. Set up template user directory in C:\USER\TEMPLATE and set up C:\NEWUSER.BAT script to create new users. Put sensible default settings in the Template dir.|
|7/23/2013||Ryan Pierce||Disabled SEM option for "windowless EDX" because it seems like we don't have it and it was inhibiting the chamber from venting. Changed resolution on EDX computer from 640x480x64k colors to 1024x768x256 colors. Now it is much less annoying with little loss of function.|
|7/21/2013||Ryan Pierce||Did full detector calibration of EDX with copper target. Also changed SEM configuration to show multi-dot EDX present, which enabled EDX presence mapping, including support for multiple channels. (Tested 3.) Noted that Isis EDX PC floppy drive failed; need replacement.|
|7/20/2013||Ryan Pierce||Added liquid nitrogen to dewar and reset liquid nitrogen temperature sensor calibration. Demonstrated that EDX is functional.|
|7/14/2013||Ryan Pierce||Changed EDX power supply voltage from 240V to 120V. Computer now recognizes detector. Customer self-test passed.|
|4/14/2013||Ryan Pierce||Ran up new filament. Filament Current 2.71A is now 1st peak. Vacuum 7.5E-5 Torr. Aligned gun and aperture. (Changed to Aperture #2.) Resolution with sputtered gold looks good.|
|4/14/2013||Ryan Pierce and Steve Finkelman||Filament blew due to vacuum loss. Replaced filament. Adjusted to 0.6mm from face of grid aperture, which is standard for longer filament life.|
|4/7/2013||Ryan Pierce||Installed new CPU fan. Finished replacing all panels except for the CPU.|
|4/?/2013||Ryan Pierce||Hacked CMOS battery from battery holder and 4x AA batteries. Had to change jumper on motherboard, per manual, to use external battery as opposed to on board battery. Replaced two of the three front panel light bulbs.|