NI InsightCM™ SDK Help
»View Product Info
This process generates new data events for data groups when various conditions trigger a data set collection, such as when the data group changes operating states or when a collection condition evaluates to true.
||Note This process writes the original waveforms acquired by the FPGA Acquisition process rather than waveforms bundled, acted on, and transferred between processes in Dataflow Script. This means any changes to the waveforms by Dataflow Script processes are discarded.
- Creates a new data event when various types of data set collection triggers occur within a particular block of data.
- Retrieves pre-trigger data and prepends it to the data in the current acquisition block.
- Updates incomplete data events with additional data when it becomes available.
- Stores channel names and properties from which other processes expect to acquire data. This is referred to as processes registering data.
Communicating with this Process
To prepare this process to write data and properties to pass through Dataflow Script, register the data. The Data Event Creator»Send Message palette contains VIs that register the following types of data: waveforms, single-point data, features and spectral bands, and speed profile data.
You can set the following properties in a software configuration definition. This process reads from the software configuration definition when it initializes.
- FileMemoryTransferMode (Default: True)—If True, the device transfers data sets to the server directly from memory rather than writing the data sets to disk before transferring them. Transferring data sets from memory helps extend the lifespan of the device by avoiding read and write operations to the hard drive.
- TdmsMemoryBlockSize (Default: 8388608)—The size, in bytes, of the chunks into which the device divides in-memory data sets from periodic acquisitions before sending them to the server.
- TdmsMemoryStreamBlockSize (Default: 1048576)—The size, in bytes, of the chunks into which the device divides in-memory data sets from stream acquisitions before sending them to the server.
Role within the Dataflow Script
When the Dataflow Script API sends a ProcessDataFlowBlock message to this process, this process requires the following inputs and produces the following outputs.
- If an ExitOpState point exists within the map, indicating the data group changed operating states, this process creates a new file and writes data of all types in the data group's map to the file.
- If a CollectFile point whose value is 1 exists in the map, this process creates a new file. In the SDK reference code, this indicates one of the following situations occurs within the current data set:
- A collection condition for the operating state evaluated true, such as a specified amount of time elapsed or the device entered a new operating state.
- An alarm instance was set or cleared for a particular alarm rule.
- This process retrieves single-point and feature values from the point value map in the data set that triggered a file collection.
- This process reads waveforms and speed profile values from the data acquired and buffered by the Accumulator process instead of the waveforms in the data set that arrives from the previous process in Dataflow Script. This means any changes to the waveforms since the Accumulator process buffered them do not appear in the data file.
- When Data Event Creator receives an acquisition block, it searches for the names of channels, speed profiles, and features that it has registered. If a channel or feature has a corresponding point in the point value map but it is not registered, this process does not write its value.
- If the acquisition block contains enough data to complete a file, this process sends a reference to the in-memory data event to the first process in the Fileflow Script to undergo post-processing.
- If the acquisition block completes a file, this process stores a reference to the file and its properties in the OpenFiles variant so this process can continue to write to it when data in subsequent acquisition blocks is available.
- Complete data events contain the following types of data:
- Waveforms from the entire acquisition duration, including pre-trigger data. Waveforms might arrive at Data Event Creator in multiple different data sets.
- Feature and spectral band values calculated from waveforms in the acquisition block that triggered the file collection. In other words, if two, four-second acquisition blocks make up the file, the analysis values are those calculated for the first acquisition block, which triggered the file.
- Single-point channel values from the most-recent data set within the acquisition block that triggered the file collection.
Customizing the Behavior of NI-Developed Processes
Post-Processing Data Events with Fileflow Script
Creating Data Events and Transferring Them to the Server
File Manager Process