Plugins

Plugin Overview

Although the base version fo Cyckei is designed to control and capture data from Keithley SourceMeters, the functionality can be extended with plugins. Plugins aim to support simultaneous data collection with additional instruments that can interface with a computer. This requires that the data capturing eventually be queried and returned by python code, but the method by which this is done is very flexible.

Installation & Configuration

Plugins are distributed independently of Cyckei as python packages. These packages are generally able to be executed independently for testing purposes, but are designed to be loaded by Cyckei for data capture.

Generally, installation involves cloning the github repository, changing to the directory and installing the python package:

To be loaded by cyckei, an entry must also be added to the plugins section of Cyckei’s config.json. Plugins should provide an example configuration and instructions on how to adjust it. The configuration for the [randomizer] plugin is included for reference.

{
  "name": "randomizer",
  "enabled": true,
  "sources": [
    {
      "port": null,
      "meta": [1, 10]
    },
    {
      "port": null,
      "meta": [11, 20]
    }
  ]
}

The configuration includes a number of reference values such as a name, whether the plugin should be enabled. It also has a list of sources that can be assigned to different channels. The mettler-ag204 plugin, for example, has the ability to interface with multiple Mettler Toldedo AG-204 scales, and declares them as several sources. The Cyckei interface then has the ability to assign different scales to individual channels for data capture. The exact parts of each source entry may depend on the individual plugin, but a port number and meta information are pretty standard. Port numbers and other information should be changed as necessary for your setup.

Running

Once configured, the different data sources exposed by plugins will be visible in the Cyckei Client.

Once available, it is as simple as selecting the source in the corresponding dropdown to assign a device to each channel. Once assigned, data from the device will be merged into the output file for that channel.

Available Plugins

The following plugins are currently known to be available. Submit a pull request to add or update entries for custom plugins.

An example of using interpreted text

Name	Purpose	Source	Version
Randomizer	Example, produces random numbers.	GitHub	0.1 Stable
Mettler AG-204	Weight from Mettler Toledo AG-104.	GitHub	0.1 Stable
Pico TC-08	Temperature from Pico TC-08 Thermocouples.	GitHub	0.1 Stable
Novus-n1050 PID	Reads data from connected Novus-n1050 PID.	GitHub	0.1 Stable

Custom Plugins

Custom plugins are simple to create, especially if there is an established method of reading device data into python already. It is recommended that you follow the scheme of the [Randomizer Plugin](https://github.com/cyclikal/cyp-randomizer).

The main component of any plugin is the PluginController class. This class is a child of Cyckei’s BaseController class which provides a number of helper functions including the essential read() method. The cyp-randomizer package includes in-line documentation to demonstrate the changes that need to be made to create a plugin for a new device. Generally most setup should be performed in the load_sources() method, and any steps to capture data should occur in the read() method. It is good practice to create some basic documentation to accompany a custom plugin, particularly if additional drivers need to be installed. Without sufficient documentation it is unlikely that plugins will be officially supported.

Another good example is the mettlerscale plugin, which gathers data from a Mettler-Toledo balance. In addition to having a read() function, this plugin utilizes a MettlerLogger object to interact with each individual scale on a different port.