This procedure/approach was developed because of problems that arose over time with the MFX slits movement. The identified issue was difficult movement by the M4 M3 slit (+X movement slit) that resulted in extreme vibrations of the entire unit, and the MDrive17 knob when turned by hand felt like it was clunking into place/there was a lot of resistance in comparison to the other turning knobs.
Consult with JJ X-Ray identified one potential issue that would not yield opening up the beamline: inside the unit each slit is mounted on a lead screw. The lead screw is connected to the motor externally to rotate by eternal motor package for rotation via a magnetic double plate system (one vacuum side and one not). These plates could be assembled situated in such a way that the gasket between them is pinched or flexed. This would lead to rubbing between the magnetic plate and cause disrupt in the movement of the slit. This would be the first way to approach the problem because it only requires venting the beamline. IF THIS IS NOT THE SOLUTION: you will have to open the entire beamline and remove the slit apparatus.
...
We began by venting the beamline according to the MFX protocol (found in MFX diamond window swap tab) with an additional PRV package at the downstream right-angle valve (preferred one has mud dauber screen and filter). Once the beamline is split open there needs to be an active purge of clean, dry nitrogen on both ends (open valve once split, before isn't necessary) and this would be the only way. Plenty of built packages are around.
For external assessment:
Vent the beamline according to protocol (keeping in mind PRVs at the source and point of use, this will require a lot of nitrogen so plan ahead)
Open the slit apparatus at the motor flange (two methods: one is remove motor first, other is leave motor on; both methods work MFX took off the motor, XCS did not)
- The motor is screwed to the flange from screws internally, so you will open up the top cover exposing the brain then find long internal screws to remove the motor itself from the flange (THIS IS IF YOU WANT TO OBSERVE EXTERNALLY for damage/warping). After that you will remove the flange from the six screws holding it into place. The attached photo is what the flange looks like
from the vacuum side with the special designed gasket removed to check for any of the interference that might have been occurring. No damage was assessed on our end, but there could be so check this area before proceeding to more drastic measures.
3. Next step after seeing no damage was to actually remove the slit from the beamline, at this point the downstream purge can be turned on (doesn't need to go on until there is a disconnection because when venting the upstream one reaches the end of the MFX beamline)
Internal assessment:
...
- when you open it up with the motor removed. You will see what they call the membrane (this is a thin flat copper gasket portion–the old design is the flat membrane with a fused o-ring on it. the fused portion has a smaller I.D. than a normal gasket)
- This is what the magnetic double plate system looks like from each end. The "clunking" that we felt as we rotated the knob of the MDrive17 motor was these circular divets bumping along into place as the plates rotate.
After seeing no damage, we decided to open up the 0 degree adapter on the side in question (NOTE: opening at the 6" portion leads to a very small window that is not large enough to observe much besides general location of the slits). This allows for internal assessment. We removed the slit from the beamline and opened up to the 8" flange.
Immediately after opening: noted a gray bar and screw sitting in the bottom of the circular chamber and the brown limit switch magnet was magnetized out of place. At the time of opening we saw no physical damage to the M3 slit we were investigating. Conclusion: the screw in the bottom of the chamber is for mounting the limit switch magnet to its place at the top of the M3 slit. The grey bar is a piece that lies on the edge of the slit itself on the internal edge (edge going towards the beam) and it was from the M1 slit (-y movement slit) on the opposite chamber side. A second screw was found on the opposite chamber side. From here we decided to remove the slit from the beamline and work under a HEPA filter.
ADDRESSING THE M1 BROKEN SLIT:
JJ X-Ray said they attached the gray bar to the slit plate using krytox brand vacuum compatible epoxy, we used torr-seal
Process: Removed the slit (just two screws holding it up and used a very tiny flat head to wear down the leftover epoxy no the slit out of the groove the gray bar sits in and blew off the shavings with clean compressed air. Once it felt like the bar sat parallel in the groove, the slit and bar were cleaned (water, acetone, alcohol, dried, blown off). Applied with torr-seal and let set under HEPA filter for 2 hour curing time. Attached once completely dry. For attachment: there is no spec on how to do this they do it all simply by eye. Advised by JJ to move the slits into position so that the two edges are as close to each other as possible and shine a light from behind. Use your eye to note if you see the plates run parallel (is there a part where light escapes through that you can see? are they overlapping at one end slightly and not the other? light is seen best when shone through a piece of blank paper). JJ said human eye measurement has the most success plus it is more important to be parallel to the slit moving in the opposite motion (M1 to M2/M3 to M4 parallel) than to be exactly level (ours were not exact flat parallel BUT are parallel to each other upon install)
ADDRESSING THE M3 DIFFICULT MOTION SLIT:
JJ had less of an idea what could be the issue outside of the gasket. Potential solution they mentioned was greasing the lead screw/linear slides.
First we attached the limit switch magnet to its place with the two found screws (they were the correct screws for this piece, still unknown by us and JJ how this could become disassembled).
Using friction between components as our first line of thought, we used UHV compatible grease to lightly grease the lead screw along the portion the slit actually travels (not the entire lead screw its unnecessary to do the whole thing) with a clean swab and manually moved the slit back and forth to lubricate the entire lead screw. After doing just the lead screw we saw no change in the tension we felt the slit had preventing motion. Each lead screw is held in place within a square inside the round chamber and as the magnet plates rotate the lead screw rotates on ball bearings held inside circular sunken holes in the square.
A small dab of grease was added to ball bearings on each side and the lead screw was rotated to spread it throughout the groove. No change in the slit motion was assessed. Each slit is held onto the lead screw via compression on a fixed component moving along a linear slide. We added grease along the rail portion of the linear slide and still assessed no change.
SOLUTION: After exhausting all potential points of friction, we stepped back to assess rotation feel. The resistance felt as though the lead screw was difficult to rotate with grease, so our final theory was that the lead screw was held too tightly within the square inner chamber. We used two clean wrenches (one held at each of the positions noted in photo below, part of the lead screw is actually flat to put a wrench on it) and loosened the nut (only 1/8 of a turn at a tiny–extremely minute movements and checking movement as going) holding the lead screw into place. This ended up eliminating the friction we felt and allowed the slit to move with comparable motion to the other three slits.
Closed up and pumped back down.
ODD NOTE: Although we did not mess with the controls, we had encoder issues after working with the slits. Take note that this might be possible.
...