Preliminary Detailed Design
All documents, pictures, and videos related to this Review can be found in the Prelim Design Documents directory.
Team Vision for Preliminary Detailed Design Phase
During this phase the team...
- Met with John Thomas, collected video data and EMG data and performed analysis to determine resting tremor frequency and amplitude.
- Continued to define project scope and focus, met with Dr. Phillips to eliminate scope creep and doubt with concept selection from previous phase.
- Defined specific issues with predicate devices so that we could design a device that would solve problems existing with the device.
- Developed a design that would minimize the effects of skin slop and utilize a set of linear brakes to mitigate tremor in all three axes of motion (flexion and extension, pronation and supination, and radial and ulnar deviation).
- Began to draft a bill of materials and test plans.
- Developed preliminary control logic.
Updated Documents from Previous Phases
Link to document: Updated Documents
Prototyping, Engineering Analysis, Simulation
Video Analysis of John Thomas
We hope to get a better understanding of the frequency and magnitude of John Thomas’s tremors in the 6 directions (Flexion, Extension, Radial, Ulnar, Supination and Pronation) by taking and analyzing video footage frame by frame. Tracking a point can provide a lot of valuable information, position, velocity, acceleration and even force if a mass or moment of inertia is determined.
First the arm will be constrained at the elbow in the block. Line up camera, aiming it down the length of the forearm such that when the wrist is flexed, the camera should be pointed between the radius and ulna. The hand and arm should then be marked with points to track. This can be done with stickers or pen marks. Recording should then started and the motions performed. Total there will be 10 tests:
- Loose Wrist, elbow straight
- Loose Wrist, elbow at 90 degrees
- Loose Wrist, elbow fully bent
- Hand straight, elbow straight
- Hand straight, elbow at 90 degrees
- Hand Straight, elbow fully bent
- Holding 5 lbs, loose wrist, elbow straight
- Holding 5 lbs, straight wrist, elbow straight
- Holding 10 lbs, loose wrist, elbow straight
- Holding 10 lbs, straight wrist, elbow straight
While some of the results were not ideal due to human error in the motions, test and data analysis, the data that was good was more than sufficient. From the select graphs above, one can see that all 3 directions of motion have a 6-7 Hz tremor. So, the frequency at which our brakes need to fire can be near constant. In addition, from this test data we can get an idea of the magnitudes. In general, the results show that Flexion/Extension Motion is most prominent. But the others are also present. The big take away for magnitude/amplitude is that the values are repetitive. Each peak is equally high as the last. This is valuable because in open loop system we can easily create an adequate response and in a closed loop system the translation between input and output will be more straightforward and predictive output is possible.
Going forward the control system design can be simplified slightly, and a few previous unknowns can now be more easily determined. While this test is not sufficient to draw any conclusions about all tremor patients, it was sufficient in giving us that knowledge that tremors are not as random as they appear. The data from this analysis will be reused in several parts of our project development including simulations and pseudo control code.
Link to document: Video Analysis Report
EMG Analysis of John Thomas
- Determine if there is a noticeable difference between EMG signal of subjects with and without tremors.
- Determine if the muscle firing in the tremor is asymmetrical.
- Determine base line tremor frequency.
Methods of Data Collection
- EMG utilized four electrode pairs as shown in diagram below (right arm), two grounding electrodes were placed on the right ankle.
- Two sets of AD Instruments’ PowerLab 26T. These directly interface with LabChart software for signal acquisition.
- Each differential voltage signal was notch filtered at 60 Hz to minimize power line noise.
- Average cyclic height of resting EMG signal was compared between healthy and tremor patient. Data acquired directly from LabChart and plots were generated in MATLAB.
- Fourier transform of tremor patient signal was computed and plotted in MATLAB to determine tremor frequency.
Resting EMG signal from subject with essential tremor. Channels 8 and 4 appear to have a periodic trend.
Average maximum cyclic height compared between electrode pair position and healthy vs. tremor subject.
FFT of EMG signal from tremor patient from electrode channel with most “visible” tremor (greatest average cyclic height).
- EMG signal from patient with tremor had a much more cyclic / periodic trend than the healthy patient. This is evidenced by (1) the appearance of the signal and (2) the average cyclic height was greater in the tremor subject at all electrode positions.
- The muscle firing appears to be asymmetrical because the average cyclic height of the EMG signal from the ulnar side of the forearm is much larger than the other side, which implies that there is likely one muscle or muscle group firing on the ulnar side of the forearm.
- The dominant frequency of the tremor was determined to be between 5 and 6 Hz. This is consistent with what was determined based on video analysis and with what is published in the literature. (The other dominant frequency in the FFT is the 60 Hz power line noise).
- Continue to design braking system that can be effective at this frequency.
- Determine feasibility of passive mitigation with current proposed design.
Link to Document: EMG Analysis Report
CAD Model of Prototype
Matlab Simulation of Breaking System using Video Data
Link to Document: Simulation Report
Feasibility: Prototyping, Analysis, Simulation
Mockup of Preliminary Design
Current Prototype Force Calculations
Link to Document: Friction Analysis
Battery Draw Calculations
Drawings, Schematics, Flow Charts, Simulations
This is a first attempt of an idea for the control logic of the mitigation device with pseudo code. There are limitations in the amount of detail due to decisions still being made on the design of the device.
- PID for adjustment of strength of braking correlating to the deviation of axis
- Will research electromagnets in future to see if the amount of current varies strength of its efficiency information
- * Strength of magnet equation - current*number of turns
- * Limitations - Heat Dissipation
- * Factors that affect performance - interference in contact to metal, current, number of coil turns, need feedback diode
- * Types - Holding electromagnets
- Different pins for amount of current given to brakes for strength/smoothing braking curve
- Multiple magnets per brake?
- Current amplifier?
Link to Document: Control Logic
Microcontroller Flow Chart
Electrical Flow Chart
Mechanical Flow Charts
Both systems are essentially the same. The goal of each is to minimize the slop as a result of interactions like those at 2., 3.5, and 6. in the figures and to optimize the braking system at 4. To better do this our system made the following changes:
- * Moved the anchor, 1., to the upper arm. The anchor there is significantly better in all 3 directions. Because it is a better anchor point, the unquantified spring and dampening constants at 2. are greater. The greater the values, the more our system approaches the ideal system.
- * By rigidly connecting the brakes to the brace (4. to 3.) we can minimize the slop between the points. The interaction, 3.5, in the old system is essentially eliminated.
- * The final critical difference is in 6. The values in our system will be higher as a result of a better glove to hand connection.
Old System Flow Chart
New System Flow Chart
Link to Document: Free Body Diagrams
Bill of Material (BOM)
Link to Document: BOM
Link to Document: Engineering Requirements
Link to Document: Risk Assessment
Design Review Materials
Plans for Next Phase
Key Questions for Next Phase
- How will we minimize heat generation within the device?
- How will we protect the user from shocking and burning?
- How will we create a stable anchor at the hand and minimize pinch points?
Goals for Next Phase
- Complete detailed design
- Finalize MSD I documentation
- Refine plan for control logic