Virtual Instrument Panel

Authored by Nikolaj Kuntner

Created on 7. June 2017

This is the Graphical User Interface & instrument driver Application Programming Interface for central laboratory device operation I wrote for the quantum computing group under Prof. Fink at IST Austria. These 15000 lines of Python are intended to be ran from an IDE. It's a Qt GUI for any collection of lab devices and works most straightforwardly with machines that can communicate via VISA, making use of NI software. It also comes with an API that provides nice state control and the possibility to track operation settings. The measurement module is tailored for superconducting QuBit measurement routines and thus fairly idiosyncratic.

Overview

VIP
├── main.py                    <─── execute that file to start the VIP
└── VIP_modules
    ├── VIP_Qwidget.py         <─── VIP class definition
    ├── widgets                <─── smaller class definitions
    ├── dictionaries           <─── contains settings (session.py, constants.py,...)
    ├── configurations
    ├── documentation
    ├── drivers
    ├── interface
    ├── scripts_for_gui
    ├── scripts_for_main
    └── external               <─── contains files not written by me (Fink group)

The primary external Python software package imported and used in the code is PyQt4, the Python binding to the Qt GUI software. An IDE which helps you with the installation might be an advantage here. Also note that unless you want to play with fonts and the Qt widgets, a large screen will help. PyQt4 is used in the main.py file, as well as the interface folders and many of the PyQt4 widgets. Note that in the present code we import widgets from two packages, PyQt4 and PySide. You may want to replace all instances of "PySide" for "PyQt4".

As a side note, the VIP Panel GUI must be opened in a Qt-Application that only lets the terminal/editor come back once the VIP-GUI is closed. A decorator that does exactly that is implemented as a wrapper function for the main function. It is found in QtGui_decorator.py the scripts_for_main folder.

Another important package is the PyVisa package, which provides a Python programming interface for the "Virtual Instrument Software Architecture" (VISA) Application Programming Interface (API), by which many instruments communicate. It is mostly used in the 'drivers' folder, defining Python classes for each instrument type. Other folders and file concern settings, infrastructure and routines. A lot of settings for the VIP itself, e.g. window size & positions, fonts, etc., are defined as constants in the constants.py file in the dictionary folder. The folder external is reserved for third party instrument drivers that suite your needs.

We run the script from the file main.py. This is the file associated with the VIP-GUI which makes use of the whole VIP software package. Note that you may need to add a few file paths to your python config or do that explicitly in main. They way the VIP package is uploaded here, it runs straightforwardly with IDEs such as Canopy.

API

The VIP is a GUI to control this settings dictionary, the vip._session, as well as to send commands to the instruments and receive and plot data. The default settings as above are hardcoded in session.py in 'dictionaries'. It is relevant to understand all the "tabs" or "session keys" (often abbreviated "tb" or "sk") and the "setting keys" (usually called "k") The settings are saved in a "session dictionary" and has the form

{session_key_1 : {settings_key_1 : value_1
                 ,settings_key_2 : value_2
                 ...
                 }
,session_key_2 : {settings_key_1 : value_1
                 ...
                 }
...
}

For example, two signal generator sources in the lab have default values as

{
...
,'SGS_33' : {'R_power_source'     : '0'
            ,'R_freq_source'      : '8'   ### A real number
            ,'F_unit_freq_source' : 'GHz'
            ,'B_output'           : 'OFF' ### A boolean value, 'ON' or 'OFF'
            ,'B_reference_osci'   : 'EXT'
            ,'B_connect'          : 'DONT'
            }
,'SGS_35' : {'R_power_source'     : '0'
            ,'R_freq_source'      : '8'
            ...
            }
...
}

In the widgets folder, the VIP class is defined as subclass of the QWidget class from the QtPy4 package that allows coding GUIs. As you can see in the code below, this file calls a main() function, which in the line

vip = VirtualInstrumentPanel()

creates a class instance and thus launches the VIP. This app is finally made accessible as a GUI via the show method vip.show(). That one above is inherited from the QWidget class, not for some custom tools. VIPinstance attributes and methods that are intended to be accessed include .set, .get, .GUI_feedback, .instruments, .result, .plot_data, ...

One could for example set the power and frequency settings for the SGS_34 via

settings = {'R_freq_source'  : "7.4"
           ,'R_power_source' : "-5"
           }
vip.set('SGS_34', settings)

or query the power setting via

p = float(vip.get('SGS_34', 'R_power_source'))
print p

or query all settings for this instrument via

print vip.get('SGS_34')

We can use the API to trigger all events that can be accessed through the GUI, for example

### writes state variables to connected instruments
import session_events as se
se.bn_connect_to_lab(vip)
se.bn_adopt_settings(vip)

which amounts to clicking the "Connect"- followed by the "Adopt settings"-button in the GUI.

Finally, in any script, one may also use methods from the drivers for connected instruments directly, for example

### Set a new current source generator frequency
f = 7402.5
SG = vip.instruments['SGS_34']
SG.set_frequency(f, 'MHz')

Almost all such operations will print an information or confirmation to the standard output.