exposure stuck at opening column valves
Added by Jean-Paul Armache over 9 years ago
Hi Anchi,
I have experienced an issue that prevented me from collecting data overnight (the collection stopped soon after I left the facility). The sample might be collected for a while, but then at some point, the app gets stuck in Exposure, writing out "opening column valves" (the rotating arrow is there all the time then), even though the column valves are not closed at all.Therefore, the collection is halted until one restarts leginon-main and returns to the session.Then, submitting the queued targets from Exposure Targeting makes the problem go away. I noticed a very similar issue in the Hole.
Has anyone witnessed it in the past and if so, found the way to make sure it doesn't happen ?
Best,
Jean-Paul
Replies (2)
RE: exposure stuck at opening column valves - Added by Anchi Cheng over 9 years ago
Opening column valves is an option that needs to be on if the scope has the tendency of closing the column valve spontaneously. Is this the case of your Polara ? The option is relatively new and only used by one other scopes as far as I know.
The logic of the program is the following: If the option is activated in your "Exposure" node with "verify opened gun valve before acquiring", AND that the column valve was closed, then you get that "opening column valves" output in the log. According to the code, it attempts to open and wait for 5 second before checking if it is indeed opened. it should "Pause" and display a "?" mark, not green arrow turning.
Since you did get "opening column valves" output log, I assume that your column valve was closed at that point. Since it was opened again when you arrive, it must be able
to complete the opening task but the program did not know that.
There are a few possibilities of what you saw, and here are my suggestions on what to do:
1. The communication to the scope was interrupted temporarily and frequent enough that you notice it in several overnight run. You should see error such as "unable to set instrument" in other places as well. Other than improving the network communication by your system admin, there is little I can do.
2. My call to open/close column valve is broken. You can verify it by clicking the "refresh" button on "Scope" node on the scope. Find the column valve display to see if it gives the right state. You can change the state there to see if column valve will indeed respond.
3. The timing for verifying it is indeed opened is too short. This step is there in case that vacuum is indeed bad, but maybe the state was not retrievable at the time. I would think that we would get an exception in the text terminal if Leginon isn't catching it, though. You can increase the wait time so that it is more appropriate for it. If it is still causing problem, you can make the assumption that it indeed opened the valve and return immediately after it opened the valve but I would not recommend doing an infinite loop to try it. If microscope closes its column valve again and again, it is protecting itself against something bad.
The option is defined in cleanBeamPath method in leginon/targetwatcher.py
Technical people in your group should be able to help you to do so if you don't feel comfortable.
The line to change waiting time:
valve_opening_wait_time_seconds = 5
The place to return if you want to skip verification is in the same method after
self.instrument.tem.ColumnValvePosition = 'open'
You should add this line and comment out the lines after.
return self.player.state()
Hope one of these will solve your problem.
RE: exposure stuck at opening column valves - Added by Jean-Paul Armache over 9 years ago
Thanks ! I will let you know how it all played out once I test it.
Best,
Jean-Paul