Serious Sensor Success! – Xan

The week began on a high note with the arrival of the Muscle Sensor Kit v3 (MSK) from Advancer Technologies, a sensor that would allow me to gather EMG data from muscle movements. Although there was a brief period at the beginning of the week when things were looking down for the idea of an EMG-based control system, everything was sorted out by Friday and a control system had been chosen! (And a rather silly human error had been discovered…)

The $50 MSK arrived on Monday night, and I was eager to get it working. On Tuesday, I began my first experiment – seeing whether or not data from the MSK could be read. I collected an Arduino Uno from an electronics parts bin, along with a breadboard, two nine-volt batteries, and some wires. After a bit of soldering and coding, I was able to gather data from the sensor, getting a number every 50ms of what the MSK was reading from the electrodes attached to my arm. Success! But the data had little variance, and I was unsure as to why…

Variance in EMG data is important for a control system like this one. Essentially, what I’m doing is sensing when the muscle in the hand recipient’s arm tenses or relaxes. When either of those things occur, an electrical signal is sent to the muscle in the arm, which the sensor reads and uses to control the prosthetic hand. If there is little variance in data, we can’t accurately determine if the muscle has actually moved or not. The data that I was getting normally varied by 1 or 2 points — and any variations in data seemingly had nothing to do with muscle tensing or relaxation, thereby making the data useless. Over 20000 samples, the highest and lowest points varied by only 5.

At first, I thought that my troubles might have had to do with electrode placement, so I experimented with different ways of placing the electrodes on my arm, and put more time into researching electrode placement. Unfortunately, the problem persisted, and I was stuck, trying to think of any ways to use the bad data that I was getting. The project was frustratingly frozen for the next few days…

On Friday, on a whim, I decided to test the two 9-volt batteries connected to the sensor. I’m sure you can see where this is going… The first battery read a voltage of 8.64V on the multimeter. Okay, not new, but not bad either. The second battery read a voltage of  about -260mV. It was not resting, stunned, or pining for the fjords – that battery was absolutely stone dead. Such a silly error was definitely embarrassing for me, but it was easily rectified by swapping out the batteries and tweaking the code once more. During the last few minutes of my Independent Research period, I connected the electrodes to my arm and was able to get lots of fluctuations in the data sent to the Arduino, all associated with muscle movements. The variance in this case was from 0 to nearly 1000, plenty to use for motor activation.

With a control system chosen, my next step will be to make a prototype for activating a motor, and to experiment a bit more with electrode placement and electrodes. After that, it will be a matter of getting a motor hooked up to the hand, and after that it will be time for the recipient to test their new EMG controlled prosthetic. The future is looking hopeful for this project!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.