As a part of my research internship at National University Singapore we had to develop a low-cost linear integrated EMG feedback circuit to control a DC-motor with its application in hand prosthesis . The circuit had to be compact and small, and at the same time efficiently filter noise from the EMG signal thus giving us a clean signal.  To help us achieve this goal active electrodes were implemented instead of the standard passive electrodes, so as to filter out significant amount of noise from the source.

What is electromyography?

The recording of the electrical activity of muscle tissue, or its representation as a visual display or audible signal, using electrodes attached to the skin or inserted into the muscle. The signals captured are known as EMG signals and are available due to the formation of electric potential on the skin surface due to muscle contraction and relaxation. An EMG is the summation of action potentials from the muscle fibers under the electrodes placed on the skin. The more muscles that fire, the greater the amount of action potentials recorded and the greater the EMG reading.

Generation of the EMG signal

The EMG is generated when a motor neuron action potential from the spinal cord arrives at a motor end plate. Its arrival causes a release of ACh (Acetylcholine) at the synaptic cleft (1) which causes a depolarization (Action Potential). This action potential electrically travels downward from the surface in a transverse tubule . This in turn causes a release of Ca++ ions, causing cross-bridge binding and the sarcomere of the muscle to contract. Magnitude of the EMG signal at the surface of the skin is between 100 to 1000 microvolts. T

Methodology

The instrumentation amplifier (INA) INA 118 is used to extract the EMG signal through EMG electrodes. The TL084 quad op amp is used to design the band pass filter, rectifier and envelop detector. The OP 177 op-amp is used as comparator to get transition pulses. Microcontroller 89S52 is used to analyze the EMG signal.

The methodology of control is a simple concept.

  • Just extract the EMG signal with help of surface electrodes and instrumentation amplifier. Filter the unwanted high frequency signal and other noise signal using band-pass filter.
  • Rectify the filtered signal and demodulate the EMG signal to get little purification using envelop detector.
  • Convert the low level signal into high level signal using comparator and it could be easy to count the transition of EMG signal using
  • Find EMG count for single action and double action of muscle by counting the transition pulses of EMG signal.
  • Run the motor in forward or reverse according to desired EMG count

The EMG count is defined as the number of positive or negative transition pulses of EMG signal for an action of muscle.

Design of signal conditioner circuit

The above figure  shows a block diagram of the signal conditioner system for extraction of EMG signal. The input stage is the dual-input instrumentation amplifier. So, two electrodes are connected to the instrumentation amplifier. The next stage is a band-pass filter and it consists of a high pass filter (HPF) and low pass filter (LPF). Then the signal is rectified with help of precision rectifier and the EMG signal demodulated with low pass filter or envelop detector. Finally the signal is connected to comparator to make compatibility with microcontroller. The cut-off frequency of HPF and LPF is 60 Hz and 1000 Hz respectively and the envelop detector cut-off frequency is 500 Hz.

Results and Discussions

The system programmed such that the motor will run in one direction if the EMG count is less than (single action of muscle) 20H or equal otherwise the motor will run in opposite direction when the EMG count is greater than 20H and less than (double action of muscle) 30H for 5 seconds and  feedback is included for motor control. The figure below  shows the EMG signal at the output of comparator and just before the microcontroller. The EMG signal is monitored with help of digital storage oscilloscope The microcontroller program is simulated with help of Keil µVision tool and microcontroller programmed or burned the microcontroller with help of Lab Tool-48 Intelligent Universal Programmer.

The simulation response of signal conditioner is done with help of Orcad family release 9.2 software. The sine wave function is selected and the frequency range is set to 300Hz and the amplitude is 0.2V. After simulating the circuit, the result waveform diagram is shown in figure below Blue color shows the comparator output, red color is envelop detector output and green is precision rectifier output. It is not possible to see the source waveform because it level is very low as compared to compared output

Repetition EMG Count for Single Action EMG Count for Double Action
1 08H 19H
2 0DH 25H
3 015H 28H
4 021H 28H
5 05H 33H

The measured EMG count is tabulated and it is shown in the table above. If we see the EMG count for single action, it varies from 05H to 21H. It has more difference and it is due to force of muscle action. When we do with less force for short time then the EMG count is less. If the force of muscle action is high then the EMG count is high.

The electromyogram (EMG) signal is the electrical physiological signal of activation of a motor unit associated with a contracting muscle and serves as a potential resource for a man-machine interface. If we see the span of the EMG count for single action and double is less. It is helpful to design a control device for a rehabilitation device like hand prosthesis

Ritwik Avaneesh, Xuang Song, Zhou Xuan, Amit Trivedi