Myoelectric
Prosthesis
Johns Hopkins Applied Physics Lab, Baltimore, MD
Alexander Sollie |Callie Wentling | Michael LoNigro | Kerry Schmidt | Elizabeth DeVito | Brian Do

Overview
Brian

Signal Collection and Processing
Myoelectric
Signals
FPGA
Microprocessor:
C2000
Brian

Mechanical Output
Base Level
• No Amputation (1:1) arm mimic
• Output to simple display system
Transradial
• Amputation “Below Elbow”
• Output to simple mechanical arm
Transhumeral
• Amputation “Above Elbow”
• Output to higher level mechanical arm
Forequarter
• Amputation of shoulder
• Output to highest level of prosthetic
Brian

Application – Medical Prosthesis
Kinematic Coupling
• Pro: Predetermine movements
• Cons: Limits prosthetic freedom
Heuristic Rules
• Pro: Reach, Move, Orient, Grasping/Releasing
• Cons: Reduces prosthetic control
Sensor Systems
• Pro: Sense environment for motion
• Cons: High variability, incorrect control
Pulse Control
• Pro: Direct control of movements
• Cons: Limited control points, higher level of dexterity
Brian

E6
E1
st
Analog
Low Pass
FIlter
Analog
RF Receiver
Analog
Analog
Active
Summer
Analog
RF Transmitter
Analog
A/D Converter
Callie

Callie

Callie

Elizabeth

Elizabeth

Elizabeth

Callie

Callie

Analog
E1
E2
Analog
Filter
Analog
RF Receiver
Analog
Instrumentational
Amplifier
Analog
RF Transmitter
Analog
A/D Converter
Callie

E Ref
ECG
PreAmplifier
Rg
Rg/2 Rg/2
INA128/INA149
Instrumentational
Amplifier
Bandpass Filter
Circuit
Low Pass w/ Gain
Full Wave Rectifier
E 1 E 2
Brian

Brian

Brian

Brian

Brian

Milestone 1
• Amplitude thresholds
• Electrode placement
• Initial Analog signal processing
• PCB Rev 1
Elizabeth

Milestone 2
• Multi Electrode system
• Amplitude Differentiation
• PCB Final Rev
Expo
• Mounting Cuff
• Consistent Signal Capture
Elizabeth

Computing
• Settled on using the FPGA with the DE2
• Experimented with C2000 but decided against it
• FPGA will connect to a board via a ribbon cable
• Most difficult part of software will be processing multiple ADC signals in parallel

ADC and Parallel processing

Control Systems and Device Driver

FPGA Details

FPGA: Input/Output Signals
•
Implement a circular buffer to hold incoming samples.
• If the amplitude doesn’t meet a certain threshold value, the samples will be overwritten.
Michael

FPGA: Input/Output Signals
Michael

FPGA: Calibration
• The previously mentioned “threshold” value will be determined by a calibration step.
• We will implement this using one of the push keys on the DE2 board.
– While the key is pressed, the values being stored in the circular buffer will be saved and the amplitude will be averaged and saved.
Michael

FPGA: Analog-to-Digital Conversion
• We would like to eventually implement 4+ channels of A/D conversion.
– The DE2 board only has a one channel ADC.
– We will use an external multi-channel ADC plugged into the D.O.U.G.L.E board.
Michael

FPGA: Additional Signal Processing
• Time permitting, we would also like to use an
FFT algorithm for more precise arm control.
– Altera MegaCore FFT algorithms would be very useful here.
• It will also be possible to control the arm based solely on signal threshold levels – no frequency spectrum analysis required!
Michael

FPGA: Output and the Motor Interface
• We’ve already successfully outputted multiple frequency square waves from the FPGA using the D.O.U.G.L.E board.
– A square wave is easily created using the FPGA and no special Digital to Analog converter is necessary.
– There are more than enough pins connected to the D.O.U.G.L.E to output different signals to multiple motors.
Michael

FPGA: Milestones
• Milestone 1
– Ability to output multiple frequencies of square wave (just one output).
– Ability to sample an incoming analog waveform.
• Milestone 2
– Implementation of a calibration step.
– Ability to sample multiple input waveforms.
– Ability to output multiple square waves of varying frequencies.
Michael

Mechanical Module
Input
• Frequency controlled square wave from
FPGA
Functionality
• Motor driver controls the stepping and supplies holding current
• Magnetic energy spins the rotor
Output
• Motor swings the forearm appropriately
Kerry

Mechanical Module
12-42V
Power
Kerry

Mechanical Module - Motor
EX -
For a 90° bicep curl:
50 Hz*1 sec*1.8°=
90°
SureStep™ Stepper Motor
• Moves the rotor discrete angles (steps)
• 1.8 degrees per step
• Holding current 2.0 A
– Holds arm in place when not in motion
– Supplies enough power to hold a small baby
Kerry

Prosthetic Arm (Higher Level Design)
Clamping motion Fore-arm twisting motion
Kerry

Milestone 1
• Bones of Arm Prototype
• Motor Driver
Milestone 2
• Working Arm Prototype
Expo
• Prosthetic with Multiple Degrees of Freedom
Kerry

Bill of Materials
Item Name / Description
Surface Electrodes
Electrode Gel (4.oz)
INA116 IC chips
Electrical Hardware (Op Amps, Resistors,
Capacitors, etc)
Hitec RCD Inc. 35990r HSR-5990TG Digital
Servo Motor
Mechanical Hardware (Motor mounts, aluminum framing, Packaging,
Prototyping)
PCB fabrication
Wireless transmitters and receivers
> RFM12B-S2 Wireless Transceivers
Altera DE0 Board and Cyclone
Microprocessor
Printing
Unit Price
400
66
175
102.5
130
12.20
20.00
10.25
150
108
Quantity
Total
1
4
1
1
1
1
4
5
2
10
Total Amount
400.00
264
175.00
102.50
130.00
61.00
40.00
102.50
150.00
432.00
Kerry
1859.00

Brian/Callie
Brian/Callie
Signals
Bipolar electrode design
Analog Filtering system
PCB Designing Board Interfacing
Elizabeth/Callie
Elizabeth/Callie
Analog Signals Board
Mechanical Driver Board
Elizabeth/Callie/Brian
Brian/Kerry
Signal PCB
Mechanical Driver Board
Computer
Independent Subsystems Independent Subsystems Part 2
Alex/Michael FPGA - Initial A/D Signal Analysis Alex/Michael
Alex/Michael FPGA - Output pulse signal to Motor Driver Alex/Michael
Basic Heuristic Laws
Calibration
Alex/Michael
Alex/Michael
Basic Sensory Systems
Advanced analog control
Elizabeth/Kerry
Elizabeth/Kerry
Mechanical
Motor Driver Interface
Prototype one: Prosthetic
Prosthetic Designing Prosthetic Interfacing
Kerry/Brian
Kerry/Brian
Aluminum Framing/Mounting Kerry/Brian
Motor Gearing System Kerry/Brian
Interfacing
Wireless Design

Schedule

Schedule (continued)

Questions?
Claudia Mitchell:
Myoelectric Arm developed by Todd Kulken
Lance Cpl. Brandon Mendez :
Myoelectric Arm, below elbow amputation
Patrick Kane, 13, is the youngest person to be fitted with a myoelectric prothetic