Actions

Stereoscan Operators Log: Difference between revisions

From HacDC Wiki

No edit summary
(november.)
 
(12 intermediate revisions by 7 users not shown)
Line 1: Line 1:
[[Category:Stereoscan]]
'''Bold text'''[[Category:Stereoscan]]


'''While repairing, maintaining, or using the Stereoscan S200 SEM, please add notes here. Newest at top.'''
'''While repairing, maintaining, or using the Stereoscan S200 SEM, please add notes here. Newest at top.'''
* November 2017. Enrique and julia fed methanol to the seals until they found a seal on the 'dented tube' that effected the pressure. pumped down to 0.34 torr. kevin and tristin wrote a program to speak to the vacuum sensor and log the data to a file. active monitoring was much more useful than the gauge at the side of the machine.
* October 2017. Enrique and Dan B. Inspected electronics and HV supply for debris. Rotated vacuum gauge to face the front. Cleaned the main door seal with ethanol. Pumped down to 4.3 Torr with the turbo spun up all the way. (It should be 10-2 Torr) There seems to be a large vacuum leak in the system. Gun seal seems OK. Next: Squirt alcohol (ethanol, isopropanol) into seals and see where the pressure goes up.
* September 2017. vacuum gauge cable has been constructed with serial feedback. we now need to experiment talking to and interpreting the data coming from the gauge.
* February 2017. Purchased and installed a more modern vacuum gauge (MKS Micropiriani 925) that reads pressure up to atmosphere and has a built-in display. This can now be used for vacuum troubleshooting.
* September, 2015. In checking the Penning vacuum gauge function, Ethan located a break in the gauge wiring; the pressure safety interlock may be keeping the filament off - a solder point failed at the high-voltage coax connector on the vacuum control board. We (Ethan, Bobby, Enrique) traced the circuit back and didn't find any voltage getting to the high-voltage transformer. Still need to figure out where exactly the voltage is failing to make it to the gauge circuitry. Ethan will trace the equivalent circuit in his board.
* August, 2015. Hunter, Ethan and Dan worked looking over schematics to see why the console reports "filament fail," despite the new, continuous filament. Ethan brought in a second 15kV electron gun power supply from another Stereoscan SEM. Replacing the gun power supply didn't change the "filament fail," status. Still need to check the vacuum status.
* July 18, 2015. Hunter, Enrique and Bobby spent all evening dealing with remaining known issues. 1) One of the pump control relays was making weak or intermittent contact. The relay was removed and found to be cracking (photo). A replacement relay (24VDC - 120VAC 15A) was found in the basement storage and soldered in with extra cables for redundancy and flexibility (the original cables are brittle). 2) The Viewsonic Composite-to-VGA converter and a suitable LCD were located and connected. 3) The video output was non-functional again. However good video could be produced by tapping the output connector on the control board. This lead to finding a broken CCTV BNC connector lead, which was soldered back to working condition. 4) Connections for the SE detector, the electron optics and the electron gun were completed following the manual and the schematics. The filament leads were tested to be continuous. 5) The stage contact alarm connector seems to have an internal short and sounds an alarm whether connected or not. It was left disconnected. 6) We began creating an index of the schematics noting which connector is listed where.
Remaining issues or suspected issues: The vacuum gauge indicator doesn't light up. This is either a problem with the gauge or indicative of a leak. We ran out of time to do an extended (30min+) vacuum test. We should also measure the gauge output manually. The system indicates "Filament Fail," despite the continuity check. This could be an interlock with the vacuum gauge, or stage current alarm. Lastly, though the video data overlay was OK, the data displayed (mag, working distance) was unstable and may indicate bad data reaching the processor.
* Saturday's working group of summakor and me ..... spent a lot of time reconfiguring the area to get back to the status of the last time I had worked on it a few weeks ago and found some minor missing things, such as a viewsonic composite/svido to VGA converter box. if found, please put with A/V gear or on the SEM. right now there's a large CRT TV on it that ... works...
* Determined that the turbo pump and roughing pump will work 100% reliably in proper wiring sequence - prior issues with starting the roughing pump and the turbo pump not running have been reduced down to one open-air relay. There are two above a torrid transformer on the column to the right of the Interlock and AUX boards. Either due to dirt or damage, it requires pressure applied from the top USING A WELL INSULATED PUSHING DEVICE THAT DOESN'T HAVE A CHANCE TO POSSIBLY ARC OR BUZZ YOU ... results in the proper operational sequence, when the column vac button is engaged, of the roughing pump pulling pressure down to turbo levels, then the turbo operationg. The Vac valve that appeared to be potentially broken works 100%. Reconnected #5 on the main PSU as well. [[User:Hunterkll|Hunterkll]] ([[User talk:Hunterkll|talk]]) 17:34, 12 May 2015 (PDT)
* Dan Barlow and 'Hunter' have determined the vacuum pump and sensor control board should be operational once a few configuration jumpers are set correctly. [[User:Mirage335|Mirage335]] ([[User talk:Mirage335|talk]]) 07:44, 21 August 2014 (PDT)
* After more thought, I've come to the conclusion that the column isn't actually missing a gasket on the specimen chamber mating surface. The column fits into a tapered bore inside the top of the chamber, and at the very tip of the column is a thick rubber fitting with a small hole for the beam to pass through. It's this very small air path that gets pumped down when the vacuum is operating, and the surface that makes the pressure seal is the rubber tip to the bottom of the taper, not the two flat disks where the column is bolted in. [[User:Plasma|Plasma]] ([[User talk:Plasma|talk]]) 21:05, 22 April 2014 (PDT)
* The replacement O-ring appears to fit on the side vent pipe of the SEM. With some help I was able to remove the column and do a cleaning of all the surfaces, save for the two detectors. It would appear we're missing a gasket for the column to chamber mounting interface, however. 4" ID, 4 3/8" OD, 1/8" thickness. I'll look to see what I can find.
Without that gasket it'd be a metal to metal vacuum fitting, and that doesn't seem right.
There's still some contamination on the bottom of the column I couldn't get off with just acetone. I'll try M.E.K. next time. [[User:Plasma|Plasma]] ([[User talk:Plasma|talk]]) 11:12, 22 April 2014 (PDT)
* Latest version of Xilinx ISE software does not support Spartan 3E board. Probably need 10.x version from http://www.xilinx.com/webpack/classics/wpclassic/ . [[User:Mirage335|Mirage335]] ([[User talk:Mirage335|talk]]) 06:43, 12 April 2014 (PDT)
* Dan Barlow commented the following on our mailing list, April 4, 2014:
#  Ran the Edwards vane pump.  Checked the oil.  Looked up the specs and maintenance requirements.
-> I recommend not doing anything else to this pump unless there is an issue.  It uses Ultragrade 19 oil, 0.6 Liter which is only about $17 plus shipping, but the current oil looks good.
#  Removed the Pfeiffer TPH190 turbo pump from the accordion flange and evaluated its mechanical condition.
http://wiki.hacdc.org/index.php/File:TurboConverted.avi
It appears the pump was intended to hang from the plinth, but the glue holding the rubber donuts to the accordion flanges has failed.  This is by far the easiest thing to fix, yay.  I replaced the turbo but its cleats are only finger tight.
#  Identified and installed the vacuum system control cabling.  Had to fix one bent pin.  The final cable from the turbo controller to the turbo itself is NOT connected, so as to avoid accidental activation.
#  Identified starting backing vacuum for the Turbo, supposedly <5x10^-1 Torr to start, <5x10^-2 torr (warning possible unit conversion failure, anyway the safety interlock should handle this) back pressure for efficient operation to provide 10^-5 in the chamber.  The thing moves 190 liters per SECOND and rotates around 30000 RPM.  It has a magnetic bearing in the top high vacuum side and a oil lubricated ceramic bearing at the roughing end.
#  Identified alternate high vacuum options in case the turbo doesn't work.  Looks like we could build an ion/titanium sublimation pump with our existing equipment and while not as fast it certainly works.


* To help keep a TCO log, I am noting that research on replacement electronics systems cost 5 hours. [[User:Mirage335|Mirage335]] ([[User talk:Mirage335|talk]]) 17:14, 3 April 2014 (PDT)
* To help keep a TCO log, I am noting that research on replacement electronics systems cost 5 hours. [[User:Mirage335|Mirage335]] ([[User talk:Mirage335|talk]]) 17:14, 3 April 2014 (PDT)

Latest revision as of 03:48, 19 November 2017

Bold text

While repairing, maintaining, or using the Stereoscan S200 SEM, please add notes here. Newest at top.

  • November 2017. Enrique and julia fed methanol to the seals until they found a seal on the 'dented tube' that effected the pressure. pumped down to 0.34 torr. kevin and tristin wrote a program to speak to the vacuum sensor and log the data to a file. active monitoring was much more useful than the gauge at the side of the machine.
  • October 2017. Enrique and Dan B. Inspected electronics and HV supply for debris. Rotated vacuum gauge to face the front. Cleaned the main door seal with ethanol. Pumped down to 4.3 Torr with the turbo spun up all the way. (It should be 10-2 Torr) There seems to be a large vacuum leak in the system. Gun seal seems OK. Next: Squirt alcohol (ethanol, isopropanol) into seals and see where the pressure goes up.
  • September 2017. vacuum gauge cable has been constructed with serial feedback. we now need to experiment talking to and interpreting the data coming from the gauge.
  • February 2017. Purchased and installed a more modern vacuum gauge (MKS Micropiriani 925) that reads pressure up to atmosphere and has a built-in display. This can now be used for vacuum troubleshooting.
  • September, 2015. In checking the Penning vacuum gauge function, Ethan located a break in the gauge wiring; the pressure safety interlock may be keeping the filament off - a solder point failed at the high-voltage coax connector on the vacuum control board. We (Ethan, Bobby, Enrique) traced the circuit back and didn't find any voltage getting to the high-voltage transformer. Still need to figure out where exactly the voltage is failing to make it to the gauge circuitry. Ethan will trace the equivalent circuit in his board.
  • August, 2015. Hunter, Ethan and Dan worked looking over schematics to see why the console reports "filament fail," despite the new, continuous filament. Ethan brought in a second 15kV electron gun power supply from another Stereoscan SEM. Replacing the gun power supply didn't change the "filament fail," status. Still need to check the vacuum status.
  • July 18, 2015. Hunter, Enrique and Bobby spent all evening dealing with remaining known issues. 1) One of the pump control relays was making weak or intermittent contact. The relay was removed and found to be cracking (photo). A replacement relay (24VDC - 120VAC 15A) was found in the basement storage and soldered in with extra cables for redundancy and flexibility (the original cables are brittle). 2) The Viewsonic Composite-to-VGA converter and a suitable LCD were located and connected. 3) The video output was non-functional again. However good video could be produced by tapping the output connector on the control board. This lead to finding a broken CCTV BNC connector lead, which was soldered back to working condition. 4) Connections for the SE detector, the electron optics and the electron gun were completed following the manual and the schematics. The filament leads were tested to be continuous. 5) The stage contact alarm connector seems to have an internal short and sounds an alarm whether connected or not. It was left disconnected. 6) We began creating an index of the schematics noting which connector is listed where.

Remaining issues or suspected issues: The vacuum gauge indicator doesn't light up. This is either a problem with the gauge or indicative of a leak. We ran out of time to do an extended (30min+) vacuum test. We should also measure the gauge output manually. The system indicates "Filament Fail," despite the continuity check. This could be an interlock with the vacuum gauge, or stage current alarm. Lastly, though the video data overlay was OK, the data displayed (mag, working distance) was unstable and may indicate bad data reaching the processor.

  • Saturday's working group of summakor and me ..... spent a lot of time reconfiguring the area to get back to the status of the last time I had worked on it a few weeks ago and found some minor missing things, such as a viewsonic composite/svido to VGA converter box. if found, please put with A/V gear or on the SEM. right now there's a large CRT TV on it that ... works...
  • Determined that the turbo pump and roughing pump will work 100% reliably in proper wiring sequence - prior issues with starting the roughing pump and the turbo pump not running have been reduced down to one open-air relay. There are two above a torrid transformer on the column to the right of the Interlock and AUX boards. Either due to dirt or damage, it requires pressure applied from the top USING A WELL INSULATED PUSHING DEVICE THAT DOESN'T HAVE A CHANCE TO POSSIBLY ARC OR BUZZ YOU ... results in the proper operational sequence, when the column vac button is engaged, of the roughing pump pulling pressure down to turbo levels, then the turbo operationg. The Vac valve that appeared to be potentially broken works 100%. Reconnected #5 on the main PSU as well. Hunterkll (talk) 17:34, 12 May 2015 (PDT)
  • Dan Barlow and 'Hunter' have determined the vacuum pump and sensor control board should be operational once a few configuration jumpers are set correctly. Mirage335 (talk) 07:44, 21 August 2014 (PDT)
  • After more thought, I've come to the conclusion that the column isn't actually missing a gasket on the specimen chamber mating surface. The column fits into a tapered bore inside the top of the chamber, and at the very tip of the column is a thick rubber fitting with a small hole for the beam to pass through. It's this very small air path that gets pumped down when the vacuum is operating, and the surface that makes the pressure seal is the rubber tip to the bottom of the taper, not the two flat disks where the column is bolted in. Plasma (talk) 21:05, 22 April 2014 (PDT)
  • The replacement O-ring appears to fit on the side vent pipe of the SEM. With some help I was able to remove the column and do a cleaning of all the surfaces, save for the two detectors. It would appear we're missing a gasket for the column to chamber mounting interface, however. 4" ID, 4 3/8" OD, 1/8" thickness. I'll look to see what I can find.

Without that gasket it'd be a metal to metal vacuum fitting, and that doesn't seem right.

There's still some contamination on the bottom of the column I couldn't get off with just acetone. I'll try M.E.K. next time. Plasma (talk) 11:12, 22 April 2014 (PDT)

  • Dan Barlow commented the following on our mailing list, April 4, 2014:
  1. Ran the Edwards vane pump. Checked the oil. Looked up the specs and maintenance requirements.

-> I recommend not doing anything else to this pump unless there is an issue. It uses Ultragrade 19 oil, 0.6 Liter which is only about $17 plus shipping, but the current oil looks good.

  1. Removed the Pfeiffer TPH190 turbo pump from the accordion flange and evaluated its mechanical condition.

http://wiki.hacdc.org/index.php/File:TurboConverted.avi It appears the pump was intended to hang from the plinth, but the glue holding the rubber donuts to the accordion flanges has failed. This is by far the easiest thing to fix, yay. I replaced the turbo but its cleats are only finger tight.

  1. Identified and installed the vacuum system control cabling. Had to fix one bent pin. The final cable from the turbo controller to the turbo itself is NOT connected, so as to avoid accidental activation.
  2. Identified starting backing vacuum for the Turbo, supposedly <5x10^-1 Torr to start, <5x10^-2 torr (warning possible unit conversion failure, anyway the safety interlock should handle this) back pressure for efficient operation to provide 10^-5 in the chamber. The thing moves 190 liters per SECOND and rotates around 30000 RPM. It has a magnetic bearing in the top high vacuum side and a oil lubricated ceramic bearing at the roughing end.
  3. Identified alternate high vacuum options in case the turbo doesn't work. Looks like we could build an ion/titanium sublimation pump with our existing equipment and while not as fast it certainly works.
  • To help keep a TCO log, I am noting that research on replacement electronics systems cost 5 hours. Mirage335 (talk) 17:14, 3 April 2014 (PDT)
  • From Tim: the red vacuum gauge sticking out of the back of the chamber is a cold cathode vacuum gauge. They're pretty reliable, but it could fail to read vacuum, indicating it needs to be cleaned. Plasma (talk) 13:07, 2 April 2014 (PDT)
  • The following steps need to be performed on the Edwards roughing pump:
    • Replace the oil
    • Inspect and clean the inlet-filter
    • Clean or replace the gas-ballast filters
    • Clean the motor fan-cover
    • Test the motor condition

Refer to pg. 38 of the user instruction manual

Plasma (talk) 23:02, 1 April 2014 (PDT)

  • Use ~~~~ to automatically sign edits with username, date, and time. Mirage335 (talk) 08:21, 1 April 2014 (PDT)
  • March 31st: Cleaned the plinth and took photographs of the vacuum components. Available here. -Dan K.
  • Oil diffusionTurbo molecular pump was hanging due to the lip falling out of the rubber suspension. Could be cleaned and probably popped back in.
  • Broken switch on roughing pump was bypassed internally. Plugging in the outlet should activate the pump. (This pump normally runs 24/7 when the machine is operational.)
  • Compressed air needed to get more dust out
  • March 30th: Moving complete! The SEM is now in the workshop
  • March 29th: Awaiting pick-up and installation (March 29th and 30th).