Leginon

h1. Pre-MSI Set-up

h2. Design Presets

To start your own MSI experiment, you must decide what presets you want to use. Presets are used to define scope and camera parameters with which images are acquired.

MSI presets can be customized to your need of data collection. By matching nodes/subnodes in MSI with presets and move types, the behavior of the node is

defined.

For easy reference, preset names in MSI has been standardized at NRAMM although there is no reason they have to be called in the way they are now. They are abbreviated to 2 letter codes to reduce the length of the filenames that contains every preset used in its family history.

Example MSI preset nomenclature:

|_.Preset name:|_.abbrev for:|_.in the context of:|

|gr|grid|em grid|

|sq|square|em grid square|

|hl|hole|quantifoil hole|

|fc|focus|focus image to be checked with fft|

|fa|focus-auto|automatic focusing|

|en|exposure-near|close-to-zero defocus exposure image in a focal pair|

|ef|exposure-far|far-from-zero defocus exposure image in a focal pair|

Example MSI preset parameters (based on 120 or 200 kV high tension, 100 um C2 aperture and 100 um objective aperture on a 4k ccd with pixel size at 1.6 angstrom at scope nominal magnification of 50,000x):

|_.Magnification:|_.Preset name:|_.Image Shift (x,y):|_.Dimension:|_.Binning:|_.Beam Coverage:|_.Exposure Time (ms):|_.Spot Size:|_.Defocus (m):|

|120|gr|Aligned|512|8|max|20|4|0.0|

|550|sq|Aligned|1024|4|1x CCD size|100|4|-2e-3|

|5000|hl|Aligned|512|8|1x CCD|20|4|-1.5e-4|

|50000|fc|0,0|512|1|<~ 1x CCD|300|4|-2e-6|

|50000|fa|0,0|1024|4|>2x CCD|50|4|-2e-6|

|50000|en|0,0|4096|1|2x CCD|170 (10e/A^2)|4|-1e-6|

|50000|ef|0,0|4096|1|2x CCD|170 (10e/A^2)|4|-2e-6|

The preset parameters in this example are chosen to give the following properties:

* *en* and *ef* are for final exposures and therefore have highest resolution and dimension.

* *en* stands for near focus exposure and therefore has lower defocus value than *ef*,the far-from-focus exposure.

* *fa* is for high resolution autofocus and therefore need to be high mag. The dimension and binning is a result of compromising speed, S/N ratio and sensitivity to

details.

* *fc* is for manual focus and ice melting and therefore more intense than *fa*. To get good FFT for checking Thon ring behavior, it is not binned. The small dimension is for increasing speed.

* *hl* is for intermediate targeting. It needs to cover the error range of modeled stage position to allow final targeting by image shift only, hence the magnification. The binning is for speed.

* *sq* is for grid square targeting. The image produced by the preset ideally need to cover most if not all of the square but with enough resolution to evaluate the content.  However, the first priority is to *choose a magnification at which no aperture used during the data collection will limit much of the illuminated area on the CCD*.  If too much of the area is obstructed, target adjustment critical to MSI operation will fail.  The coverage of the grid square can always be achieved by choosing multiple targets on it.

* *gr* is for producing grid atlas. It is selected based on the size number of tiles needed to cover the whole grid in an acceptable time and the amount of distortion common to very low mag imaging.  Objective aperture can be removed and a larger C2 aperture used while using this preset to acquire grid atlas.  It is never used after the atlas is acquired.

In designing your own preset, follow the above properties using a mag that gives the required coverage with your scope, camera, and grid mesh. If you have a 2k or 1k camera, binning may not be as necessary as for a 4k camera whose data acquisition time is 10-30 sec without binning.

h2. Another Recent Preset Design Example (Aiming for better than 5 A resolution, although not yet proven)

Tecnai F20 200 kV high tension, Gun Lens 3 extraction voltage 4300, 50 um C2 aperture  and 100 um objective aperture (except when obtaining grid atlas).

Camera is a 4k ccd with pixel size at 1.6 angstrom at scope nominal magnification of 50,000x):

|_.Magnification:|_.Preset name:|_.Image Shift (x,y):|_.Dimension:|_.Binning:|_.Beam Coverage:|_.Exposure Time (ms):|_.Dose (e/A^2):|_.Spot Size:|_.Defocus (m):|_.C2 (um):|

|120|gr|Aligned|512|8|max|20|n/a|4|0.0|100|

|1700|sq|Aligned|1024|4|1x CCD size|10|n/a|4|-2e-4|50|

|5000|hl|Aligned|1024|4|1x CCD|7|0.004|4|-1.5e-4|50|

|100000|fc|0,0|512|2|<~ 1x CCD|200|164|4|-2.9e-7|50|

|100000|fa|0,0|1024|4|~4x CCD|50|2.5|4|-1.5e-6|50|

|100000|en|0,0|4096|1|1.3 um on the specimen|407|20|4|(-1e-6 to -2e-6)|50|

|29000|ed|0,0|3kx4k|1|1.3 um on the specimen|407|20|4|(-1e-6 to -2e-6)|50|

* Note: This experiment used direct detection camera for final exposure (ed preset). en preset was not used in the actual experiment. It is here just for reference and to show what we would have used if recorded on CCD based on our GroEL study.  During process, the final images for en would have been binned by 2 based on our experience so the en preset could have being collected at 2048 binned by 2.

* Note: no defocal pair was aquired in this experiment.  Therefore, no ef preset.

h2. Calibrations

If this is the first time calibrations have ever been completed for Leginon, complete all the calibrations listed in [[Leginon "Calibrations" Application|the chapter on calibrations]]. Calibrations are extremely important to operate the TEM in a consistent manner. As long as the calibrations are stable, later users do not need to perform them. Most calibrations are HT and magnification dependent. Therefore, new calibration may be required if your presets use different HT and/or magnification from all previous users.

______

[[Summary of MSI applications|< Summary of MSI applications]] | [[Initial MSI application preferences|Initial MSI application preferences >]]

______