Skip to content

Frequently Asked Questions

The Knight Board is an eight-channel biopotential acquisition board for EEG, EMG, ECG, EOG, and other ExG signals. The signal type depends on electrode placement, reference arrangement, gain, and the software processing applied to the recording.

How many channels and sample rates are available?

Section titled “How many channels and sample rates are available?”

The board provides eight 16-bit ExG channels. Supported sample rates are 125, 250, and 500 samples per second, with 125 SPS as the default. Each channel has an independently configurable gain of 1, 2, 3, 4, 6, 8, or 12.

See Knight Board Specs for the complete hardware table.

Does the board need a battery or external power supply?

Section titled “Does the board need a battery or external power supply?”

No. The Knight Board is powered through its USB-C connection and streams data over the same cable. Use a USB cable that supports data and power; a charge-only cable can power a device without creating a usable serial port.

The EXG Visualizer is available for Windows, macOS, and Linux. BrainFlow and LSL can also be used from supported programming environments on those platforms. Serial-port names differ by operating system:

  • Windows: COM3, COM4, and similar
  • Linux: /dev/ttyACM0, /dev/ttyUSB0, and similar
  • macOS: /dev/cu.*

The Knight Board appears as a USB serial device. In most environments the operating system supplies the required serial support. BrainFlow includes native definitions for the standard and IMU Knight Boards, so a custom packet parser is not required when using BrainFlow.

If no serial port appears after connecting the board, follow Board Is Not Detected.

Use the named BoardIds value rather than copying an integer from an example:

BoardIds.NEUROPAWN_KNIGHT_BOARD # standard Knight Board
BoardIds.NEUROPAWN_KNIGHT_BOARD_IMU # Knight Board with motion sensors

The IMU ID exposes accelerometer, gyroscope, and magnetometer channels in addition to EEG. See Board Setup for a full connection example.

Why is the signal empty after BrainFlow connects?

Section titled “Why is the signal empty after BrainFlow connects?”

Connecting starts the BrainFlow session, but EEG channels are disabled in software by default. After start_stream(), wait for the board to settle and send a chon_{channel}_{gain} command for every channel you need. Leave pauses between commands so the firmware has time to apply them.

The physical channel switch and the software command control different parts of the signal path; both must be configured correctly. Follow the Knight Board Command Set startup sequence.

With a switch away from the header, that channel uses the common reference input. With the switch toward the header, its negative input is routed to the individual N header pin. For a typical common-reference EEG montage, move each used channel’s switch away from the header. Switch unused channels off to reduce noise and crosstalk.

The individual negative inputs also allow bipolar or mixed biosignal montages. See Knight Board Specs before wiring a custom reference arrangement.

COMM is the common reference input. For the standard EEG setup, connect it to the ear-clip reference. RLD is the driven reference output and can use the second ear-clip connection to reduce common-mode interference.

RLD is optional, but it often improves recordings in electrically noisy environments. When using it, add the active channels to the RLD network with rldadd_{channel} commands.

No. Dry spike electrodes are designed to make contact through hair without gel. Part the hair so the spike tips reach the scalp, then tighten only enough to maintain stable, comfortable contact.

Paste or gel can be used with spike electrodes to lower contact impedance. Gold cup electrodes require paste or gel. Wet contact can increase signal amplitude, so start with a lower gain such as 4 or 6 and check for clipping before increasing it.

What is the difference between spike, comfort, and gold cup electrodes?

Section titled “What is the difference between spike, comfort, and gold cup electrodes?”
  • Dry spike electrodes record through hair and can be used without gel.
  • Comfort electrodes support and balance the headset; they do not record.
  • Gold cup electrodes are wet recording electrodes used with conductive paste or gel.

Read Headset and Electrodes for placement and gain guidance.

Gain 12 is the normal starting point for dry EEG electrodes. Lower the gain if the trace reaches its limits or appears clipped, especially when using paste, gel, EMG, ECG, or another larger-amplitude signal.

When using BrainFlow, the gain in BrainFlowInputParams.other_info must match the gain sent in each chon_ command. Otherwise, BrainFlow will use the wrong scale when converting ADC counts to microvolts.

Can I use my own electrodes or headset accessories?

Section titled “Can I use my own electrodes or headset accessories?”

Yes. The board accepts compatible passive and active biopotential electrodes, provided they are wired to the correct positive, negative, reference, and RLD connections for the intended montage.

Headset-mounted electrodes and custom accessories use a fully modeled external M16×2, class 6g, right-hand ISO metric thread. The headset node contains the mating internal cavity formed from the negative of that thread profile. See Creating Custom Accessories before designing or printing a part.

Can I record different biosignals at the same time?

Section titled “Can I record different biosignals at the same time?”

Yes, with an appropriate montage. The individual N1N8 inputs allow channels to use separate negative electrodes instead of the shared COMM reference. This can support combinations such as EEG and EOG, but each signal may need a different electrode placement and gain. Verify that the selected gain does not saturate any channel.

The EXG Visualizer can monitor signals and export CSV data. BrainFlow returns a two-dimensional array containing EEG and metadata rows, while LSL can publish a live stream to tools such as LabRecorder.

Use get_current_board_data() for a non-destructive live window. Use get_board_data() when you intentionally want to retrieve and clear the BrainFlow ring buffer. See Knight Board Data Format and the BrainFlow API Reference.

Does the board include wireless streaming?

Section titled “Does the board include wireless streaming?”

The Knight Board connects directly over USB-C. It does not require a wireless adapter for normal operation. For a wireless project, connect it by USB to a Wi-Fi-enabled host such as a Raspberry Pi and forward the resulting BrainFlow or LSL data over the network.

Can I modify the software and build my own BCI?

Section titled “Can I modify the software and build my own BCI?”

Yes. NeuroPawn provides BrainFlow and LSL integrations plus complete SSVEP, motor-intent, and P300 tutorials. The public repositories include example code that can be adapted for custom experiments.

The documentation and tutorials are intended for biosignal learning, research, prototyping, and development. They do not provide a medical diagnosis or replace clinical equipment, professional interpretation, or emergency care. Do not make health or safety decisions from an unvalidated recording or tutorial model.

Check Troubleshooting first and include the board variant, operating system, serial port, software version, wiring or montage, gain, and the exact error when asking for help.