IV - Maintenance and adjustment

[ Software Settings | Rate meter | Voltage meters | k-space calibration | R-space calibration | Stop Watch | Logfile || Main Help Index ]

1. The Software Settings form

[ General | Forms | File I/O | Substrate | DAQ | LEED | Ranges ]

In the Software Settings window, you can customize several global program settings such as colours, paths and specific parameters of your LEED system and data acquisition hardware. For all tabs there are the "Apply" and "Default" buttons either for applying the settings in the currently opened tab or for loading the default settings.

The general software settings tab ("General")

The general software settings tabs contains some global settings of WinSPA that are not related to SPA-LEED at all. You can (should!) set the installation and data paths and your standard printer. A change of the colour scheme of the scan data displays is optional, but may be necessary depending on your colour scheme (Microsoft XP Luna, Microsoft Vista Aero).

Note: The log file is written into the folder specified by the data path. Earlier releases of WinSPA (including V1.0) used the program path for the logfile.

The graphical user interface settings tab ("Forms")

In the "Forms" tab you can customize several details of the visual appearance of WinSPA's graphical user interface.

The file I/O settings tab ("File I/O")

In the "File I/O" tab you find the settings for data import and export.

WinSPA uses the free data format NetCDF (network Common Data Format). NetCDF is a machine-independent format for representing scientific data. This format is widely used in the scientific community. NetCDF files can be handled by numerous software products such has IDL (native) and Matlab (requires a toolbox). C, Fortran, Java, etc. libraries are also available to read/write netcdf data files.

In particular, NetCDF data is

• Self-Describing:
  A netCDF file includes information about the data it contains.
• Architecture-independent:
  A netCDF file is represented in a form that can be accessed by computers with different ways of storing integers, characters, and floating-point numbers.
• Sharable:
  One writer and multiple readers may simultaneously access the same netCDF file.
• Direct-access:
  A small subset of a large dataset may be accessed efficiently, without first reading through all the preceding data.

All information on how to deal with netcdf format can be retrieved at Unidata. Free NetCDF browsers for viewing data are also available (NCbrowse and Panoply).

Note: In the NetCDF tab you can see whether the dynamic link library (DLL) for NetCDF file access has been found or not. You should check this after the installation of WinSPA. If the DLL has not been located by WinSPA, re-check the presence of the netcdf.dll and click on the try again button.

In the Substrate Settings tab you can change the parameters describing the surface of your current sample. This is necessary for the internal calculation of reciprocal space units.

The substrate settings tab ("Substrate")

In the "substrate" tab you can enter the name, step height and lattice constant of the sample you are using. These values are used if you choose %BZ as k_par units and/or the scattering phase S as k_perp units.

Note: Each time you use a user defined substrate, it will be automatically added to the presets list (for WinSPA V1.06 and higher).
If you frequently use other samples than the predefined ones, you may also add them to the samples list in WinSPA's config file with the standard Windows text editor.

Note: If you use a sample which has neither quadratic nor hexagonal surface symmetry then %BZ are not an appropriate k_par unit, since the k_x and k_y are then different. Use either the deflection voltages or inverse Anstroms for those samples.

The DAQ settings tab ("DAQ")

In the DAQ tab in the Software Settings window you can define the basic settings of your SPA-LEED system. The values of deflection and emitter high voltage amplifier gains have to be correct for proper operation of WinSPA with your hardware.

Note: The timing strategy of the counter can heavily influence your measured results. Please make sure you are aware of the following aspects before you change settings of the timing strategy.
The high voltages for deflection and acceleration do not change ideally synchronous to the respective software commands. After a dead time due to software execution times, hardware driver communication and loading of the values to the D/A converters, the high voltages will show an exponential creep from last stable value to the newly set one, see below.

On modern PC systems the dead time are negligible in comparison to counter gate time of 0.1ms and above. The time constants of the high voltages may, however, be significant and are therefore taken into account via the "counter timing" settings.
The finite speed of the high voltages is inherent, since HV amplifiers are usually designed to have a high output impedance (safety and power considerations) and the cables and deflection plates are electrically a capacitance. Together, output impedance and load capacitance, they form a first order low-pass circuit (RC filter). The deflection voltage may not have to drive a "large" capacitance of deflection plates, but usually several control voltages inside an electron gun are coupled via high value resistors (anode, lens, acceleration potential) in order to derive the needed voltages from only one reference voltage. The effect is comparable to the one described for the deflection voltages. For the sake of stable and low noise high voltages additional low pass filter elements may have been added by the designer of your SPA-LEED system. As a rule of thumb, the deflection voltages of an Omicron SPA-LEED are very fast, the old Leybold systems suffer under much slower low-pass-filtered deflection voltage amplifiers. The electron energy settling is comparably slow in both systems.
The time constants for the exponential creep is entered in the LEED system tab, since they are typical for the used hardware. The way how to deal with finite HV voltage speeds is defined in the DAQ tab. Please keep this in mind.
Since reaching the exact high voltages level would take infinitely long, the idea is to define an "acceptable" deviation between set deflection or acceleration voltage and the "as is" value which is internally calculated from time constant and delay time. "Acceptable" means of course always a compromise between speed and precision. The timing strategies basically involve to

• accept an absolute error (for example 0.05 eV for th electron energy is definitely negligible and below the converter noise of the Burr-Brown 12bit D/A converter), which leads to zero delay times during a scan line and possibly very large delay times between scan lines
• accept a relative error (e.g., reaching 90% of the total specified step response), which follows the usual timing definition for all industrial data acquisition systems, see t63, t90 times; this leads to a constant delay time independent of voltage step, and
• accept the minimum (fast) or maximum (precise) of each of the above described values, or
• ignore the behavior of the high voltages (old SPA-LEED softwares, fastest)

Note: Depending on your SPA-LEED hardware, the chosen scan type (bidirectional, direct retrace) could lead to artefacts in your acquired data. Please make sure you are fully aware of the meaning of these settings before you change them.
The finite response times of the deflection voltages as described above are the reason for the two scan types you can choose from. Both are meant for reducing artfacts caused by the finite seed of the high voltages even without delay times.
In a bidirectional scan (valid for 2D, RSM, SEM) neighboring lines are scanned in alternating directions. This minimizes the voltage differences between the last point of one scan lines and the first point of the next one and therefore reduces the absolute error of the deflection voltages at the beginning of each scan line. The drawback, however, is that the small relative deflection voltage errors may now become visible since the errors in adjacent lines have opposite signs: Comb-like distortions of diffraction spots may appear for small gate times and slow deflection amplifiers.
A direct-retrace scan (also valid for 2D, RSM, SEM) the idea is implemented to use the scan idle times for letting the deflection voltages settle: The first points of adjacent scan lines correspond to a very similar pair of x/y-deflection voltage values. Therefore the deflection voltages are set to the values of the first scan point after a scan line is finished. The time for displaying the scan data on the graphical user interface is then "used" for the settlement of the deflection voltages. No specific artefacts are caused from this setting, it is therefore the standard preset.
The "usual" scan mode of the SPA-LEED as implemented in older software packages, where each scan line is finished, the data are displayed and then the next scan line is started, could cause artefacts especially at small counter gate times and slow deflection amplifiers: It may occur that the counting at the first point of the next scan line is already started while the deflection voltages are still somewhere between last point of old scan line and first point of new one. This leads to fake intensity at the first one or few scan pixels of each scan line, since the counter acquires intensity during the creep of the deflection voltages.
The image below shows schematically the behavior of the different scan types.

The combinaton of direct retrace and bidirectional is allowed but does not make sense: The first scan type is based on the idea of using idle times for large scan voltage changes while the second one minimizes the scan voltage changes. The combination of both simply combines the artefacts of both without any benefit. Therefore a warning is displayed in case you activate both check boxes in the DAQ settings tab.
Hint: Take a 2D scan with a small gate time (1 ms or less) with the "bidirectional scan" setting active, and observe the inter-digital-like shifts of the scan lines. If these shifts are small, you may use the "birectional scan" setting and not delay timing (see next tab) for fastest possible scans. If the shifts are large, you should introduce delays to allows for high voltage settling.

Note: The beam park position is used to protect your channeltron during idle times, since the latter's lifetime at the approx. 10⁶ counts per second from a specular spot is limited to only a few hours.
Check always that your park position setting does not accidentally coincide with a diffraction spot position.

The SPA-LEED system settings tab ("LEED")

In the LEED system tab you can set several (hardware) parameters of your SPA-LEED system, including geometry and beam deflection values.

Note: Some scan types, especially the RSM scan, rely on a "perfect" alignment of the sample. This means that the (00) spot is centered in the diffraction images solely mechanically, the sample is placed in the specified distance in front of the instrument, and no electromagnetic fields are present.
This ideal case is hard to achieve. Especially, a perfect mechanical alignment requires to angles of rotation of the sample holder (rotation and tilt) as well as a time-consuming iterative alignment if the sample holder does to rotate around the point of incidence of the electron beam onto the sample surface. Therefore WinSPA allows a correction of angular misalignments, which cause a linear image shift upon changes of the electron energy.

Note: If you are using an Omicron third-generation SPA-LEED with a sample distance of 15 mm, then the preset settings are correct. In other cases, please take the necessary time for measuring and calculating (angles) as well as calibrating (sensitivities). Refer to the section "standard operating procedures" for further details.

Note: The R space sensitivity may be dramatically different depending on how you realized the SEM-like mode of operation: You may either change the polarity of one of the two octopole deflectors in your instrument or simply ground one of them, preferably of course the one at the sample end of the instrument, not the one at the emitter/channeltron end.

The tab for global ranges settings ("Ranges")

In the "Ranges" tab you can modify the allowed input value ranges.

The values of many graphical user interface input elements of WinSPA are checked to be within a defined range of allowed values. Even though this patronizes the user, severe problems can be avoided this way, such as

• Memory overflow due to giant scan sizes.
• System lock due to long gate times.
• Emitter damage due to overvoltage arcs between emitter and anode.
• ...and many more.

There may be, however, situations where the user does not want the software to try to be more intelligent than the user. For example, you may want to make a 2D scan with 10,000 by 100 pixels, knowing that this costs less memory than the default 4000 by 4000 pixels. Or you might want to view a 100 by 100 pixels 2D scan in 20 times magnification, knowing that this would not be a good idea for a 4000 by 4000 pixels 2D scan since the internal graphics memory might not be sufficient.

In those cases only you might change the range settings of WinSPA. Please think twice why the preset ranges are not sufficient for your application and whether any problems could be caused by exceeding the predefined ranges. Also, please set the default values immediately after you have finished the scans for which you needed the extended settings.

Note: Since the standard settings are considered to be correct for all "normal" applications of SPA-LEED, your changed settings are not saved in the config.dat file - they are only valid until closing WinSPA.

XY offset voltages

For adjustment purposes and for a correct RSM scan alignment you can set offsets on the deflection voltage in the x- and y-directions. During "normal" operation the offsets should be set to zero, since a good mechanical alignment of your sample is independent of the electron energy and therefore better than electrical shifts.

Manual energy setting ("Manual E")

In some situations, you may want to set the electron energy manually and directly , for example to find an out-of-phase condition, mainly when you are looking at the phosphor screen of your SPA-LEED. In case the SPA-LEED supply is too far away to reach the energy control while looking at the phosphor screen, "manual energy" allows you to use mouse and keyboard (extension cables may be useful) to change the electron energy quickly and observe the effect immediately.

Note: WinSPA crashes if the manual energy setting is used in parallel to the execution of any scan. This is caused by the way the energy setting and scan execution are handled internally. This may be changed sometime in the future, but in fact there is really no good reason for tinkering with the electron energy during scanning!

2. The rate meter

In the ratemeter and the voltage meters windows you can find some display elements for checking and controlling the I/O data of WinSPA, such as count rate, deflection voltages, electron energy and more.
All those values are displayed via a callback function from the SpaDAQ.dll. That means that they come from a very deep abstraction layer of the software. You can therefore be pretty sure that these values are correct (depending on the voltage gain of your deflection and emitter high voltage amplifier). The drawback of this callback approach is a greatly reduced system performance (bidirectional communication with data acquisition library) and lots of problems with a thread-safe and lock-proof access of the Delphi VCL (visual control library).
In order to avoid those problems, the meters are now designed (from WinSPA V1.05 on) to work asynchronously from the value changes. This means, the meters are udated a little bit more often than 10 times per second with the latest values. As known from any time discretized system, this can lead to sampling artefacts (moire, beats).

3. The voltage meters

The voltage meters allow you to monitor the output voltages of WinSPA.

Note: The values are updated when and only when SpaDAQ.dll updates the output values. In order to minimize hardware access, the values are only written to the hardware if and only if a value has changed. You might therefore see E=000.00eV quite some time after switching on the "show voltages", even though the electron energy is not 0 eV. This value is only updated after current electron energy is changed the first time after switching on the display (the same holds for the deflection voltages).

4. The k-space calibration

The k-space calibration form is a tool that makes it very easy to perform a calibration of the reciprocal space of your SPA-LEED system.

Note: The k-space calibration is one of the standard operating procedures. Please refer to the respective section of this help.

5. The R-space calibration

The R-space calibration form is a tool that makes it very easy to perform a calibration of the real space of your SPA-LEED system.

Note: The R-space calibration is one of the standard operating procedures. Please refer to the respective section of this help.

6. The Image Shift calculation

In the mage shift calculation form you can enter the position of a diffraction spot (the specular spot in most cases) at different electron energies. A linear function can then be fitted to the positions, and their parameters can be copied to the software settings form.

7. The Stop watch

The stop watch is a small but very helpful tool for efficient measurements. It basically defines a local time in in-situ experiments for the start of the measurement, the beginning of the deposition etc..

Note: Not being meant as a replacement for a well-documented lab book, of course, the "log this:" allows you to store short text lines along with a precise time stamp into the text log file of WinSPA. The text information is usually something like "evaporator shutter opened", "sample temp. raised to 500K" or "refocused".

8. The log file

WinSPA saves all scan events (start/stop of scans), stopwatch actions and text comments ("log this:") into the log file. This file can therefore be a valuable help for your data analysis, since even complex measurements can be sorted and analyzed afterwards.

The start times of the scans are also written into each scan file, as well as the local stopwatch time. This information is therefore redundantly saved. Your comments and all start/stop information of scans you do not save to disc are only written into the logfile.