sarmadtec-neuroport 0.1.0

Sarmadtec NeuroPort driver for Fascin8/Ultim8 neuro headstages with faker mode for offline testing.

Sarmadtec NeuroPort Driver (pip package)

Python driver for Fascin8/Ultim8 neuro headstages with optional fake signal generation for offline testing. The code is packaged as sarmadtec_neuroport so it can be installed via pip and reused outside this GUI app.

Install locally

• Editable/development install from this repo: pip install -e .
• One-off install from a checkout or tarball: pip install .
• Requires Python 3.8+ and pyserial (pulled in automatically).

Quickstart (faker mode)

from time import sleep
from sarmadtec_neuroport import Device

def on_data(samples):
    if samples:
        first = samples[0]
        print(f"{len(samples)} samples; seq={first[0]} channels={len(first)-1}")

device = Device(
    debug_mode=True,               # print driver logs
    faker=True,                    # built-in signal generator instead of hardware
    data_ready_callback=on_data,   # called when new samples arrive
    data_chunk=0,                  # 0=flush everything each callback
)

device.sampling_rate = 500        # 250/500/1000/2000 Hz
device.gain = 24                  # ADS1299 gain setting for primary channels
device.exgain = 24                # gain for extra channels
device.leadoffMode = True         # enable lead-off detection

device.configure_settings()       # push settings to device (no-op for faker)
device.start()                    # begin streaming
sleep(3)
device.stop()
device.shutdown()                 # cleanly close worker threads/serial port

Run python examples/basic_usage.py --faker --duration 3 for a ready-made script. Drop --faker to talk to real hardware over USB/serial.

Real hardware notes

• Plug in the Fascin8/Ultim8 so the OS exposes a CDC/serial port (Mac: /dev/tty.*, Windows: COMx).
• The driver auto-scans for descriptors containing “USB Serial Device” or “STM32 Virtual ComPort”.
• You can observe connection changes by registering connection_state_callback:

  Device(connection_state_callback=lambda connected, port, reason: print(connected, port, reason))
  

• Call reset_device() to force a reconnect/re-handshake without recreating the object.

API cheat sheet

• Device(debug_mode=False, data_ready_callback=None, settings_written_callback=None, data_chunk=0, faker=False, faker_data=None, connection_state_callback=None, connection_check_interval=1.0)
  • Instantiation starts internal I/O + watchdog threads and kicks off connect_async().
  • faker=True bypasses serial I/O and emits synthetic samples; optionally replay faker_data (list of channel-value lists).
  • data_chunk=0 flushes all buffered samples per callback; otherwise batches that many.
• Core methods:
  • connect_async(force=False) / reset_device() — handshake with the headstage.
  • configure_settings() — send current sampling/gain/lead-off/channel flags.
  • start() / stop() — start/stop streaming; parsed samples surface via callbacks or get_data().
  • set_signal_generator_mode(enabled, faker_data=None) — toggle faker mode on an existing instance.
  • set_data_ready_callback(cb) / set_connection_state_callback(cb) — register runtime callbacks.
  • shutdown() — close serial, stop worker threads/watchdog, and reset state.

Samples are parsed as Python lists: [sequence, ch1, ch2, ...] (Fascin8 produces 21 primary + 3 extra channels; Ultim8 produces 8 primary).

Build and publish a wheel/sdist

python -m pip install --upgrade build twine
python -m build            # creates dist/*.whl and dist/*.tar.gz
# twine upload dist/*      # push to PyPI when ready

That is all that is needed to ship the driver as a pip package. Update the version in pyproject.toml and neuroport_driver/__init__.py before publishing new releases.