Abstract:
Skeletal muscles are the key source of our body motion. To perform any activity, messages are transmitted from brain to specific muscle through motor neurons. Muscles are tied up to the bones via tough cords of tissues and these cords are called tendons. As the muscle contracts, it pulls on the tendon, and resultantly cause the bone to move. Contraction of muscle is due to nerve impulse stimulation. Because of stimulation, exchange of ions across the muscle fiber occurred, which result in generation of small electrical current, which combined for a specific motor unit, is known as the Motor Unit Action Potential (MUAP). Combination of all electrical signals generated from all of the MUAPs in a detected area is called myoelectric signal. This signal is known as Electromyogram (EMG).
Humans with amputations due to mishaps (wars/accidents) or inborn absence, are forced to use prosthesis. Two types of prosthesis are being used worldwide, passive and active but latter one is more expensive but considered as substitute to natural limbs to some extent. To control myoelectric prosthesis, SEMG signals are picked up through surface electrodes (dry/wet). Multiple active/passive SEMG sensors have been developed and are being used worldwide to control it. Each type has its own merits/demerits. To get noise free EMG signal, this thesis research has focused to develop active SEMG sensor, which could control myoelectric prosthesis effectively. A novel simulink model is developed, which mimics various elements of an active SEMG sensor. By using this model, low/high/notch filters are designed and optimized, which are used in signal conditioning of raw signal acquired from SEMG sensor. Finally, on the basis of successful simulation results, instrumentation of surface EMG sensor (38mmx24mmx 7mm) is designed and fabricated using Altium Designer 14. Pure silver electrodes (17mmx5mmx1mm) are used as dry electrodes and are directly connected to Pre-amplifier (INA118). Results of the Simulink model and developed SEMG sensor have shown close resemblance, therefore, the model will be used for further research in domain of electromyogram.
