As shown in Figure 1, the hardware requirements for
this project are minimal. If you’ve been experimenting in
robotics for any period of time, the only thing you’ll likely
need to buy is a Myo Ware EMG sensor ($38, Amazon)
shown in Photo 2, and a supply of adhesive electrodes
($25 for 50, Amazon). Any Arduino-compatible
microcontroller and servo gripper or other servo-controlled
device will do.
You can buy a dozen adhesive electrodes for about $10
from Amazon, but if you plan to do much experimentation,
you’ll go through them in a few days and then you’ll have
to wait for delivery of additional electrodes to continue your
In addition to the base Myo Ware sensor, I ordered an
external lead daughterboard. Together, with a set of two
foot EMG leads, the add-on board was supposed to allow
me to keep the main Myo Ware sensor on my desk.
However, I couldn’t get the extension leads to work.
Nonetheless, the external lead board was useful in
providing additional connection points for the Arduino and
my digital oscilloscope.
I used a 9 VDC battery to power the circuit instead of a
line-powered supply out of an abundance of caution. In the
rare event of a power supply short or lightning strike with
the EMG electrodes in place on my arm or chest, I could
possibly receive a potentially fatal shock. A simple optical
isolator between the EMG sensor output and the Arduino
analog input is an option if you really want to use a wall
wart or other line-powered DC source.
Figure 2 shows the overall servo gripper circuit. Not
counting power and ground, there are only four
connections to the Arduino. The EMG sensor output is
applied to analog input A0. The LED output is D12, and the
servo PWM output is through D7. The momentary contact
button, S1, is connected to pin D10.
Note that this is a closed-loop real time system that
employs a translational reality interface (TRI). That is, the
user both controls and monitors the servo gripper by virtue
of a momentary contact switch mounted on the jaw of the
gripper and a signal LED that indicates jaw closure. Because
of this architecture, the user could conceivably control a
remote servo gripper or any other device by contracting
their bicep or other skeletal muscle, and monitor their
success or effort with the LED or other indicator. Of course,
the system components can be used locally, as in the local
control of an artificial limb.
By Bryan Bergeron
SERVO 04.2017 43
Post comments on this section and find any associated files and/or downloads at
Figure 1. Hardware
Photo 2. MyoWare EMG sensor, showing button
connectors for adhesive electrode.
Closing the Loop