Design and Validation of a Customizable Electromyographic Sensor Module for the DLR Research Wristband

Abstract

This master’s thesis presents the design and validation of a customizable surface electromyography sensor module tailored for integration into the DLR research wristband, a modular wearable platform aimed at advancing human-machine interfaces. Surface electromyography enables the non-invasive detection of electrical activity produced by skeletal muscles, providing valuable insights into user intent for controlling assistive devices and interactive systems. The thesis addresses critical challenges in sEMG signal acquisition, such as minimizing power line interference, optimizing signal-to-noise ratio, and improving spatial and temporal resolution through electrode and circuit design. A systematic approach was employed involving circuit simulation, circuit and electrode design, and sensor module validation. Four circuit topologies, single differential with three or six electrodes, double differential with four electrodes, and monopolar with six electrodes, were designed and analyzed. Simulations evaluate gain, commonmode rejection ratio, and frequency response. Sensor modules were validated in comparison to a commercial sEMG system. Experiments employed an isometric contraction protocol on the biceps brachii across multiple load levels to assess signal quality, frequency characteristics, and the ability to detect propagating motor unit action potentials. Results demonstrated that the designed sensor modules achieved comparable signalto-noise ratios to the commercial reference system, although steel electrodes presented challenges related to increased common-mode noise that could be mitigated by notch filtering. The results revealed that single and double differential acquisition modes significantly affect the amplitude and frequency characteristics of measured EMG signals, which must be considered when interpreting sEMG data or designing sensor systems. The integration of the sensor module into the DLR research wristband was successfully validated through a multi-subject study, confirming consistent sEMG acquisition under various loading conditions and highlighting battery-powered operation benefits. The findings substantiate the feasibility of a customizable, modular sEMG solution for wearable applications.