Design and Metrological Characterization of an Open-Source, Low-Cost sEMG Data Acquisition Instrument for Neuromuscular Signal Monitoring

This dataset provides the complete open-source design files and firmware required to replicate the ultra-low-cost (<€50) surface electromyography (sEMG) data acquisition instrument described in the study "Design and Metrological Characterization of an Open-Source, Low-Cost sEMG Data Acquisition Instrument for Neuromuscular Signal Monitoring". The instrument is based on a modular hardware-software stack utilizing an ESP8266 microcontroller, a 12-bit ADS1015 ADC, and a SEN0240 induction sensor. The provided files ensure full research transparency and technical reproducibility. Content of the repository: Hardware Schematics: Detailed electrical wiring and circuit diagrams in PDF format. Gerber Files: Production files for the custom PCB (IoT-ready version). Firmware: The MicroPython binary and source code for the ESP8266, optimized for stable 1000 SPS data acquisition via I2C (400 kHz). DSP & Processing Tools: Documentation of the 4th-order Butterworth and Notch filter parameters used to achieve a Signal-to-Noise Ratio (SNR) of 29.72 dB and 0.80 Pearson correlation with clinical standards. This project aims to democratize access to high-fidelity neuromuscular sensing for applications in assistive robotics, prosthetics, and human-computer interaction (HCI). This repository provides the software stack for a high-fidelity surface electromyography (sEMG) acquisition system designed for benchmarking low-cost hardware against clinical-grade standards. The system utilizes an ESP8266 microcontroller and an ADS1015 12-bit ADC (connected to SEN0240 sEMG vía I2C) to achieve a deterministic sampling rate of 1000 SPS, ensuring strict temporal and spectral alignment with reference databases such as NinaPro DB1. This method ensures a stable 1000 Hz sampling frequency by bypassing the Flash memory write latency of the ESP8266. The accompanying offline Python engine performs the complete digital signal processing pipeline, including 4th-order Butterworth band-pass filtering (20–450 Hz), 50 Hz digital notch filtering, full-wave rectification, linear envelope extraction via a 3 Hz leaky integrator, and polyphase resampling to 100 Hz. Hardware integration requires an ESP8266 clocked at 160 MHz connected to an ADS1015 via I2C on GPIO 0 (SCL) and GPIO 2 (SDA). The ADC address must be set to 0x48 by grounding the ADDR pin. The DFRobot SEN0240 sensor is connected to channels A0 and A1 in differential mode. Repository Structure: -Firmware: MicroPython code for the ESP8266 (acquisition program): EMG_ADS1015_1000SPS_80s_v4.py -High-speed binary acquisition script via i2c with ADS1015 -Processing: Python scripts for PC-based data analysis (to obtain statistics peak, MAV, RMS, WL and visualization): plotter_data(10Rx8seg)_binary_mv_stats_2.py -Pearson Correlation Analysis: -Pearson Correlation Analysis: Pearson_correlation_validator_with_OFFSET_SAMPLES.py -Classification: Implements a Random Forest Classifier sEMG_Gesture_Recognition.py -SNR Analysis: sEMG_SNR_calculator.py