ironbci

At a glance

Channels	8 biosignal inputs, plus reference and bias
Converter	TI ADS1299, 24-bit, simultaneous sampling
Sample rate	250 SPS
Signals	EEG, EMG, ECG, EOG
Microcontroller	STM32
Wireless	Bluetooth Low Energy, appears as EAREEG (some units IronBCI)
Electrodes	Dry or wet Ag/AgCl, 2.54 mm header
Battery	LiPo, 200 mAh recommended, 2.0 mm JST, charged over Micro USB
Software	PiEEG Cloud (browser), Android app, PiEEG Server, Python SDK, JavaScript SDK

How it works

 Electrodes 1–8 ─┐
 Reference (ear) ┼─► ADC board ─► BLE board ─► Bluetooth LE
 Bias (ear) ─────┘   (ADS1299)    (STM32)        │
                         ▲                       ├ cloud.pieeg.com
       LiPo battery ─► Power board               ├ Android app
                                                 ├ PiEEG Server
                                                 └ Python / JS SDK

Quick start

1. Connect the battery to the JST connector under the board. Check the polarity first.
2. Charge: Switch 1 ON, Switch 2 OFF, plug in Micro USB. LED 1 = charging, LED 2 = full. Unplug the USB when done.
3. Connect 8 electrodes, a reference ear clip and a bias ear clip.
4. Switch on: both switches ON. The board starts advertising over Bluetooth.
5. Connect with any of these (no pairing in system settings needed):

Option	How
Browser (fastest)	Open cloud.pieeg.com in Chrome or Edge → PiEEG Devices → IronBCI
Android	Install mobile_app/mobile_sdk.apk, choose 8 ch in settings
PiEEG Server	pip install "pieeg-server[ironbci]" then pieeg-server --device ironbci8
Python	pip install bleak matplotlib numpy then python SDK/BLE_General/3.Read_data_graph.py

When connected, the BLE LED turns on and 8 channels stream at 250 samples per second.

Full walkthrough: docs.pieeg.com → IronBCI-8 → How to Start

Battery, switches and LEDs

Mode	Switch 1 (battery)	Switch 2 (data)
Recording	ON	ON
Charging	ON	OFF

LED	Meaning
LED 1	Charging
LED 2	Battery fully charged
BLE LED	Bluetooth connection active

Wiring

Input	Electrode	Purpose
Channels 1–8	EEG electrodes in a cap or headset (10-20 system)	Signal
Reference	Ear clip on one earlobe	Common 0 µV point for all channels
Bias	Ear clip on the other earlobe	Cancels common-mode noise (mains hum)

Suggested positions (from the IronBCI paper): F7, Fz, F8, C3, C4, T5, Pz, T6.

Software

Python SDK

Scripts built on bleak in SDK/BLE_General:

Script	What it does
1.Find_Device.py	Scan and print the device’s Bluetooth address
2.Read_data.py	Connect and print values in µV
3.Read_data_graph.py	Live graph of all 8 channels
4.Read_gpaph_Autoscale.py	Live graph with automatic scaling

pip install bleak matplotlib numpy scipy
cd SDK/BLE_General
python 1.Find_Device.py
python 3.Read_data_graph.py

BLE protocol: data characteristic 0000fe42-8e22-4541-9d4c-21edae82ed19 (notify), 8 channels × 3 bytes (24-bit big-endian two’s complement) per sample. Details: Specification.

PiEEG Server

pip install "pieeg-server[ironbci]"
pieeg-server --device ironbci8                                   # scan and connect
pieeg-server --device ironbci8 --ble-address AA:BB:CC:DD:EE:FF   # skip the scan

Dashboard at http://localhost:1617, plus recording, WebSocket, LSL and VRChat OSC.

Android app

Install mobile_app/mobile_sdk.apk (MIT licence), choose 8 ch in settings, press Save to record and Stop to finish, then share the file. The Flutter source is in mobile_app/EEG-Flutter-App-master.zip. Based on BodyPress.

JavaScript SDK

const eeg = new PiEEG();
await eeg.connectBLE();          // IronBCI-8 is the default; call from a click handler

See the JavaScript SDK docs.

All options: Software docs

Example signals

Recorded with dry Ag/AgCl electrodes, no gel, at 250 SPS.

EEG: alpha rhythm (eyes closed and open) · dataset

EEG: chewing and blinking artifacts

Projects

• Real-time emotion detection with ChatGPT: Chat_GPT/Chat_GPT_Real_time.py streams EEG, computes alpha power every second and asks an OpenAI model to label it. Guide: Connect Brain to ChatGPT.
• 3D-printable headset: ironbci_3D_EEG_Printable_Headset.
• Your own mobile app: start from the Flutter source in mobile_app.

Videos

Device introduction	Mobile SDK app test	How to program a BCI

What’s in this repository

Folder	Content
SDK/BLE_General	Python BLE SDK: find, read, live graphs
mobile_app	Android app (.apk) and Flutter source
dataset	Alpha test, raw and 8–12 Hz filtered (dry electrodes, 250 SPS)
Scripts_for_work_with_dataset	Band-pass filter and plot recordings
Chat_GPT	Real-time emotion detection with an OpenAI model
3D_headset	Link to the 3D-printable headset
Supplementary files	Images used in the original README

Documentation

Page
IronBCI-8 Intro	Overview, comparison with PiEEG-8
What’s in the Box	What you need besides the board
How to Start	Battery, charging, switches, first connection
Electrodes & Battery	Wiring, montage, switches and LEDs
Software	PiEEG Cloud, Android, PiEEG Server, Python, JavaScript
Signals & Projects	Example recordings, dataset, ChatGPT project
Specification	Hardware, BLE data format, µV conversion
Troubleshooting	Battery, Bluetooth and signal fixes
Safety	Battery-only operation and LiPo care

Citation

If you use IronBCI-8 in research, please cite:

I. Rakhmatulin. IronBCI: A Low-Cost, Open-Source BCI Platform with Mobile SDK for
Rapid Neurotech Prototyping. Preprints 2025, 2025071198.
DOI: 10.20944/preprints202507.1198.v1

Contact

• 🌐 pieeg.com
• 📖 docs.pieeg.com
• 💬 Discord
• ✉️ [email protected]