Helping Hand

Oct 2018 - Dec 2018

30.002 Circuits and Electronics

30.001 Structures and Materials

As part of both the Circuits and Electronics as well as Materials and Structure modules, we are tasked to build a product that is able to tackle a problem in the real world.

We choose to target the healthcare sector, specifically patients that lost hand functions due to stroke. Patients that suffer from stroke has a high change (~87%) of paralysis of hand or upper limbs. To recover the loss of hand functions, they have to undergo rehabilitation therapy that is often held in periodical sessions.

However, outside of those sessions, the stroke patients do not have the chance to exercise their grip strength as it is relatively weak and unreliable on day-to-day operation.

Our product is an exoskeleton for the hand that uses Electromyography (EMG) to assist gripping actions of the user.

There are 3 electrodes involved - 2 electrodes are paste onto the biceps and 1 onto the elbow. When the user flexes their biceps, it will generate a signal and activate a linear actuator that is connected to the exoskeleton. This will, in turn, push the exoskeleton finger and provide additional grip strength to the user.

Since stroke patients that have loss of hand functions have weaker forearm strength, they will be using their biceps to activate the assistance instead.

For this product, we use purely circuitry without any microcontroller.

Firstly, the potential generated by flexing of the biceps are fed through the instrumental amplifier and amplify the differences, as well as rejecting the common noise.

Secondly, it is fed to an active high-pass filter with a cut-off frequency of 2hz.

Next, it undergoes full-wave rectification by passing the signal through a precision amplifier.

Subsequently, it goes through an active low pass filter that cuts off 110hz as well as generates a signal envelope.

Following which it is pumped to an inverting amplifier to invert the signal back to positive.

Finally, it is fed to a comparator to decide if the signal is high enough (flexing). If it is, it will activate the linear actuator forward, vice versa. The threshold voltage is controlled by a potentiometer so that it can be adjusted for different user's preferences.

Special thanks to Akib and Zhi Xin for designing the exoskeleton structure and the poster. Without you guys, it would not have been fun!

In this project, I was in charge of designing the schematic and build theing circuit. Through this project, I was able to learn so much more about filters and op-amps usage then theories from classes. There were so many times the past schematic failed to work which was somewhat frustrating. I am just happy that the final design works out well.

In the spirit of the upcoming Christmas and just for fun, we decided to make a video for it.

Solution

Read up many papers regarding EMG
Many trial and error
Use Falstad to test
Use Arduino as digital Oscilloscope.

Unable to design a suitable schematic that is able to read the flexing of muscle accurately

Problems faced

Op-Amps (Filters/Rectifier/Comparator)
Signal processing
Schematic design`
Geometry design