Table of Contents

NeuropixelsV1e Headstage Configuration

The following excerpt from the NeuropixelsV1e Headstage example workflow demonstrates how to configure your NeuropixelsV1e Headstage and Breakout Board in Bonsai. This process comprises of the following steps:

~/workflows/hardware/np1e/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.

Configuring the NeuropixelsV1e Headstage

The HeadstageNeuropixelsV1e operator is used to configure the Neuropixels V1e Headstage; this can enable streaming of electrophysiology data from a Neuropixels 1.0 probe and orientation data from a Bno055 IMU. This is accomplished in the example workflow by leaving all of the HeadstageNeuropixelsV1e properties set to their default values.

Default values for the headstage are:

  • Enabling the first 384 electrodes for streaming (electrodes 0 through 383)
    • This is also known as the Bank A Channel Preset
  • Setting AP Gain to 1000x
  • Setting LFP gain to 50x
  • Enabling the Spike Filter
  • Setting the Reference to External
Important

The workflow will not run unless gain calibration and ADC calibration files are provided. Click the HeadstageNeuropixelsV1e operator, expand NeuropixelsV1e in the properties pane, then choose the appropriate files by selecting either GainCalibrationFile or AdcCalibrationFile and clicking the ... button.

Tip

For additional details on how to manually configure the headstage, such as enabling specific electrodes for recording, or modify AP / LFP gain, check out the NeuropixelsV1e 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 NeuropixelsV1e Headstage example workflow, most collected data is from the NeuropixelsV1e Headstage. The rate of data being produced by the hardware will be ~20.6 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 20.6 MB/s the hardware will produce 8192 bytes every ~400 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 NeuropixelsV1e Headstage that can be configured through the ConfigureHeadstageNeuropixelsV1e operator, visit the NeuropixelsV1e Headstage Overview.

Edit this page