background image


Electromyography (EMG) is a technique for detecting and recording electrical potentials generated during muscle contraction. In particular, EMG surface potentials (signals which travel along the surface of the skin) typically range from 50 uV to 10 mV, with a frequency range of 0 to 500 Hz. EMG signals are very random in nature and are therefore extremely difficult to both detect and reproduce. The goal of the EMG signal generator is to take a data set representing previously captured EMG signals, scale them using an analog input (potentiometer), and then output both the modified signal and its gain. The resultant signal can then be input into a prosthetic device.

The software and hardware components of our project work together to provide the best possible resolution while precisely adjusting the gain of our output signal. The software running on our microcontroller parses the complete signal and scales it up before sending it to the hardware component. By scaling the signal up in software first, we take advantage of the full resolution of our digital to analog converter. Once the signal is transferred to the analog hardware section of our project, we can adjust its amplitude very precisely. This precise amplitude adjustment is best accomplished in analog hardware, which we can control using our software. Through this project we have improved the existing Bioelectric Signal Simulator by the addition of an adjustable gain (attenuation) control and digital ''system gain'' display. There have also been additions to the project firmware so as to enable user programmable ''preamplifier gain'' parameters for each signal as well as code to maximize the resolution of any signal through the system's D/A converter, regardless of initial signal level.

These additions will allow the system to be used in the calibration of myoelectric devices such as EMG detection and capture systems and myoelectric prostheses. In the immediate future, the system will undergo a round of rigorous testing to ensure it's performance and then be used at CanAssist, at the University of Victoria, in the qualification, modeling, and calibration of a myoelectric prosthetic arm.

In the future, this bio-electric signal generator could be used as a research tool in designing the next generation of prosthetics. It could also be used as an add-on to optimize existing systems or as a testing tool for all types of bio-electric signals. An electrical system that adds amplitude scaling to signals has potential for use in a variety of signal processing applications.