Table of Contents

NeuropixelsV2e Headstage Configuration

The following excerpt from the NeuropixelsV2e Headstage example workflow demonstrates how to configure your NeuropixelsV2e Headstage and Breakout Board in Bonsai. This process comprises of the following steps:

~/workflows/hardware/np2e/configuration.bonsai workflow

NOTE

To learn more about the top-level configuration motif in every workflow involving ONIX hardware, visit the Configuration Chain Tutorial.

Creating an Acquisition Context

The CreateContext operator creates a ContextTask that defines the device driver and index where the hardware exists. The Driver property is set to "riffa" which is the name of the PCIe device used by ONIX. In this case, the Index property is set to 0 because there is only a single ONIX system. If a second system is used on the same computer, a second CreateContext operator would be required in its own configuration chain, with its Index property set to 1.

Note

The NeuropixelsV2eBeta Headstage functions nearly identically to the NeuropixelsV2e Headstage. Simply replace ConfigureHeadstageNeuropixelsV2e with ConfigureHeadstageNeuropixelsV2eBeta.

Configuring the NeuropixelsV2e headstage

The HeadstageNeuropixelsV2e operator is set to configure the NeuropixelsV2e Headstage; this can enable streaming of electrophysiology data from a Neuropixels 2.0 probe and orientation data from a Bno055 IMU. This is accomplished in the NeuropixelsV2e Headstage example workflow by leaving all of the HeadstageNeuropixelsV2e properties set to their default values.

Default values for the headstage are:

  • Enabling the first 384 electrodes of the first shank for streaming (shank 0, electrodes 0 through 383)
    • This is also known as the Shank 0 Bank A Channel Preset
  • Setting Reference to External
  • Setting Invert Polarity to True
Important

The workflow will not run unless gain correction files are provided. Click the HeadstageNeuropixelsV2e operator, expand NeuropixelsV2e in the property pane, then choose the appropriate files by selecting either GainCalibrationFileA or GainCalibrationFileB and clicking the ... button. If only one probe is plugged in, only one file is required.

Tip

For additional details on how to manually configure the headstage, such as enabling specific electrodes for recording, check out the NeuropixelsV2e GUI page.

When the workflow is started, the current time (based on Coordinated Universal Time) is saved, along with global hardware parameters governing data acquisition. This is accomplished using a TimeStamp operator to capture the computer's wall clock time. The timestamp is saved along with ContextTask's properties (e.g. AcquisitionClockHz, BlockReadSize, BlockWriteSize) to a csv file (start-time_<filecount>.csv) when the workflow is started.

Starting Acquisition

After starting a workflow, the StartAcquisition operator begins data acquisition with the hardware that has been configured. In the NeuropixelsV2e Headstage example workflow, most collected data is from the NeuropixelsV2e Headstage. The rate of data being produced by the hardware will be ~18.1 MB/s. The ReadSize property is set to 8192 bytes, meaning data collection will wait until 8192 bytes of data have been produced by the hardware. At 18.1 MB/s the hardware will produce 8192 bytes every ~440 μs. This is a hard bound on the latency of the system. If lower latencies were required, the hardware would need to produce data more quickly or the ReadSize property value would need to be reduced.

The WriteSize property is set to 2048 bytes. This determines the amount of memory that is preallocated for temporarily holding data before it is sent to hardware. It is less critical to performance unless the rate that data be written to the hardware is comparable to the rate that the hardware produces data, which is not a common scenario.

NOTE

For an overview of the devices on the NeuropixelsV2e Headstage that can be configured through the ConfigureHeadstageNeuropixelsV2e operator, visit the NeuropixelsV2e Headstage Overview.

Edit this page