Leginon

The settings look normal.

There are three possibilities:

1. The shutter is wired so that the post-specimen shutter is shutter 2. You can also set both shutters to close when idle, so that you don't have to figure out which is which.

• unplug and replug in the connector to the microscope shutters. There are labels on the panel for what is what. This sometimes reset it.
• restart the camera controller according Gatan's instruction.

3. You've left the camera in a retracted state for too long while at the high magnification. This is a problem Gatan should solve, but it is a tricky one. When the camera is retracted, the beam is not blanked even if we did set it to do so in the configuration. It had to let go the blanking in case there is another camera below, but that is in conflict of low-dose principle.

Hello Anchi,

Thank you very much for your help. I have changed the setting as you suggested:
1) Set the shutter 2 to close when idle
2) I have restarted the K2 summit (the analyzer, the digitizer and the computer)
3) During my data collection, I keep the camera inserted and did not let it idle for a long time.

The images look better but the radiation damage still exists. My sample got burned after about ~10 frames (about 20 e/A^2 in total, which should not burn my sample). I checked our methods again and I think that two more possible reasons,

1. Our pre-set cycle is gr->sq->hl->fc->tomo->fm. When the tomography is collected, the Leginon goes through the cycle from tomo->fm->gr->sq->hl then hl->fc->tomo. I have used Gatan DigitalMicrograph software to acquire a image on our sample. Then I run this cycle once without screen down nor camera viewing. But after the cycle, I used Gatan DM to collect the image on the same site. I noticed that strong radiation damage had occurred. So my guess is that when the Leginon went through the pre-set cycle, the electron beam keeps on the sample. Is this assumption possible?
2. I remember that when we used K2 camera in counted mode in Scripps last November, you have mentioned that the beam for data collection should spread large enough to cover the entire phosphor screen and the beam would be very dim to reach the right dose at about 10 e/pixel/s. And the measure exposure time was about 1.6 seconds. However, when I try the similar settings in our FEI TF20, the dose displayed in Gatan only reaches about 2 e/pixel/s. And the measure exposure time is also longer at about 10 seconds. So I have to focus my beam to a relatively small beam size to reach electron dose at 6 e/pixel/s and exposure time at 4.0 seconds. I know that each electron microscope has different settings compared with others. But I just wanna to make sure that this radiation damage is not due to the TEM.

Thanks a lot!
-Long

Long,

The size of the beam that will give you 10 e/pixel/s does depend on high tension/ gun lens /spot size /condenser aperture size as well as the state of your electron gun. The sizes we were using at Scripps was not the entire phosphor screen but more likely the size of the 4 cm diameter circle. We chose the spot size/ gun lens so that we are in that range so that not too much area was exposed but not imaged. I added the beam diameter in Preset Design for K2 Camera. Even if your exposure meter or DM value not quite accurate, you should know it is intense or not easily. The intensity used for counted mode is really hard to see on the main viewing screen when the room is lit. If your beam has been too strong, the camera would have retracted itself.

Regarding the radiation damage:

If the shutter is not closed while idle as the configuration in DM claimed, it would burn even if you are not changing preset.
I suppose this would have been obvious: If you use low dose kit of the microscope to emunate preset change (i.e., mag and beam size change), and check with DM afterward, there would be no burn?

Also, on our scope, we often see the beam blank comes on during the last stage of the main screen lifting. Do you see that ?

There is a setting in Presets Manager called "Beam Blank during preset change", you can try activating it. We have not had to use this ourselves but it can be used as extra precaution. This activates beam blank from the microscope side. However, the external input from the camera is meant to override it. Using it in combination with "Cycle magnification only" means that camera parameters are not set during cycling, can therefore its shutter state would not have changed even if there is a problem with our DM scripting call.

The most direct test of shuttering signaling is the voltage output of the two connectors to the shutters at K2 Digitizer box. Get someone to help you with doing some voltage or resistance measurement on the output from K2 controller with a multimeter. I can't be sure which is which, but it should flip between 0 and ~3.5 V between shutter off and on, the typical digital on/off. Read the value while idle and the value during an exposure will tell you if it is sending signal. Do this test with DM image acquisition as well as pyscope K2 test script. If DM can flip the shutter but Leginon can not (through pyscope), then we have a problem of changing DM Scripting behavior with this new version of DM. If even DM can not flip the shutter signal, then you should contact Gatan. If both flipping the shutters properly, then the problem would be on the microscope side.

Hello Anchi,

Thank you very much for your detailed reply.

I have just prepared several new grid and tried them on Leginon with both "Beam Blank during preset change" and "Cycle magnification only" checked. And the initial images we collected are not damaged. However, the following tilting images after the 5th tilting image are burned. Based on our previous experience, our virus sample could bear at least 100 e/A^2

So here is my guess: maybe the shutter system of our K2 camera could not work during the tomography data collection. It's supposed that the shutter would be closed when the cryo-holder is tilting and moving. It is possible that our new DM software could not communicate with Leginon client, which results in open shuttle during the tomography data collection. And this open shuttle would finally burn our gird.

So tomorrow Kelly and I plan to test the shuttering signaling on the K2 digitizer box as you suggested. I have attached the figure on the back of K2 Digitizer Box. Should we just pull out the cable from the shut 1 and 2 ports, connect the ports with probes from multimeter and then record the signal changes when the shutters are supposed to be off or on?

Thanks a lot for your help again. I will let you know the results as soon as we collected.

Best,
Long

Yes, that is the idea.

Because the tilting does not involve preset change. It is likely that the shutter configuration was not set by DM gui as we normally assume. Since the beam blank signal used in "Beam Blank during preset change" is issued in a different way. Let me see if I can figure out how to set that on our side instead.

Hello Anchi,

We have figured out that the sample got burned due to an error during the K2 camera installation. The shutter cable from K2 digitizer box is supposed to connect with pre-specimen shutter of the TEM but it was connected to the post-specimen shutter by mistake. Now the problem is solved. Thank you very much for the help and I am sorry to disturb you so much about this issue.

Thanks again,
Long

I am glad it is resolved now. One puzzle: you did test setting shutter 2 also to close when idle. Therefore that should have given you undamaged result ?

Hello Anchi,

Actually only 1 shutter from K2 digitizer box is connected to the TEM. The other shutter cable is not connected to anything else. I did not know whether it is designed by Gatan or it is a mistake during the installation. I will contact them later.

Thanks,
Long