Setup

Station Configuration

The station configuration contains information about the test setup itself, including any instrumentation connected to it. The simplest way to import a configuration is via JSON file, though any dict with the correct formatting will work.

For loading a file station_config = StationConfig.read_json(path=Path('station_config.json')). For creating a StationConfig from a dict, use station_config = StationConfig({...}).

Key	type	Description
station_id	str	A single unique identifier of the station itself.
paths	dict	locations to read and save data. Two keys are supported; base, whose value is the parent directory for all other directories listed, and data, whose value is the name of the directory where all raw data is saved.
instruments	dict	The keys of this dictionary are the instance names of the instruments. These instance names must match those in the recipe for the instrument to be used. The value of these keys is another dictionary that contains the instrument class and information about connecting to its com port.

Recipe

A recipe represents the tests to run over a device under test. The simplest way create a recipe is by writing a JSON file, though any dict with the correct formatting can be loaded.

For loading a file recipe = Recipe.read_json(path=Path('recipe.json')). For creating a Recipe from a dict, use recipe = Recipe({...}).

"instruments": ["smu", "ps"],

This element lists instrument instance names that are needed to execute this recipe. Every instrument needed by the all the tests in the recipe need to be included. If the recipe requires both an SMU and a power supply, for example, the the entry might look as above. These instance names must match those in the Station Configuration (make this a link to the relevant section above). The station configuration determines which SMU to use and how to connect to it. Not all instruments present in a station need to be listed, only those needed for this specific recipe execution, however, if an instrument instance does not exist, testing will fail during initialization.

Individual tests are defined in the test block, an example is shown below:

"tests": [
  [
    "abc",
    {
      "class": "VirtualTest",
      "init": {},
      "acquire": {},
      "analysis": {
        "report_headings": ["Virtual IV"]
      }
    }
  ],
  [
    "bcd",
    {
      "class": "NewTest",
      "init": {},
      "acquire": {},
      "analysis": {
          "report_headings": ["New Virtual IV"]
      }
    }
  ]
]

The tests value is of type list[list[str, dict]]. The tests are executed in the order listed. Each element of the tests list contains two parts, a str, which is the test instance name, and must be unique (TODO: CHECK) and a dictionary with the following elements:

Key	type	Description
class	str	A valid test class registered within the framework.
init	dict	A kwargs for the initialization of the test class. The recipe can pass additional information to the test this way. If nothing is needed, the value should be an empty dict {}
acquire	dict	A kwargs for the data acquisition portion of the test execution. If the TestExec is run in analysis_only=True mode, the data acquisition portion will not execute so these options will not be used. If nothing is needed, the value should be an empty dict {}
analysis	dict	A kwargs for the data analysis portion of the test execution. If nothing is needed, the value should be an empty dict {}

Adding a Test

To add a new test, you need to create a class which inherits from autosweep.tests.abs_test.AbsTest. Remember that the TestExec only calls the __init__, run_acquire, and run_analysis methods during recipe execution.

The script that runs the TestExec must also register the test with AutoSweep before running the recipe by calling the function register_classes with the argument being the module that contains the new test.

Adding an Instrument

To add a new instrument, you need to create a class which inherits from autosweep.instruments.abs_instr.AbsInstrument.