Final project
So I started searching for prosthetics hand designs to be inspired and Mr. Neil has referenced an association for prosthetics called E-NABLE that was very useful. I found another website “https://kumu.io/”. They have listed every open-source design like a tree of design and it was very fascinating and very inspiring, and I found a lot of resources of a deficient source of designs. but I didn't find any of them to explain if there is a rule for designing a prosthetic hand or an example of how you can create your own device.
So at the begging, I thought it will be an easy and challenging experience but I was wrong in every way, it just needs a lot of design practices and a lot of trying and error but with the coronavirus pandemic and getting sick most of my plans have been destroyed X”D, and I will explain step by step what I have completed and every problem I have faced in the process.
First for my final project plan I have a lot of ideas to implement so to make everything easy I decided to make“flow chart” blocks for every step I need to do to make functional hands so I have the full explaining in “Applications and Implications” but as I have explained because coronavirus I didn’t have this match time to do everything so I will make a smaller flowchart to explain what I really will do in the final project.
EMG
So first How our muscles work?
starts out from your brain the signal goes down through your spinal cord from your brain to the main motor neuron health your muscle and that motor neuron when it gets activated by your brain signal fires an electronic signal down the length of your muscle and what that signal does is it creates polarization and depolarization in your muscle known as the action potential and that action potential spreads down your muscle and activates the motor neuron axons in your muscle and all the sarcomeres along the strand of your muscles are activated by the signal the myosin in the sarcomeres are activated they grab hold of the actin strands on either side and pull them together and thus your muscles contract.
My first temptation
So I will use the EMG to sense the muscles contraction, my main core of the project is the EMG sensor so on my first try I did make a small amplifier circuit to connect the gate and ground to my arm to see if it will detect any electrical activity so I hooked my circuit to the oscilloscope to see if it will detect and thing and I failed but I think I discovered another way to make a touch sensor X”””D
Anyway I searched for a simple EMG sensor Cricut and I found a lot of examples but my main problem was in the component availability in my country soI took another approach first is to understand the main concept of the EMG Sensor.
As you can see the circuit concept of the EMG Is to use a Differential amplifier with 3 electrodes placed on the body, two in close proximity 2 cm minimum, and the other on the elbow as a reference.
Next for the Filter, The High pass filter is necessary to remove all contribution from <20Hz frequencies to reduce the most common corruptions of your signal – movements of electrodes and the low-pass filter attenuates all frequencies above 500Hz
So for the amplifier, as I said I didn’t found any decent replacement In the local market, So I decided to look If there any amplifier module or sensor has a chip fit the requirements that I need lucky I found an LM324 Differential amplifier It in ultrasonic sensor “hy-srf05”
So I found a multiple EMG circuit use LM324 components value was very different, so I calculated the High pass filter to see if there any Cricut will be similar to my calculation, I found this great website to accurately calculate the components value.
A similar circuit to my calculation was found on CHARLES' LABS Website so I used some of his idea as a starting point
So I started to make my first version of the PCB I used through-hole components for my first prototype to make this easy at first.
Next for testing, I used the oscilloscope to see if the circuit works and what kind of signal the PCB generates when Trager my Muscles.
So surprisingly the PCB worked from the first time It was generating peaks between either 0 or 5V. If it is set the pot to too low
Now because life is short so I don’t have to discuss with you the last four failed versions X””””D
But i will till what i was trying to do is to develop the Emg sensor so I but it directly into my arm Muscles without using the electrode patch
And i separated the ICL7660 from the emg main circuit because i will make it in a spirit pcb because i discovered that the emg circuit was only take between 0.05 and 0.09 amp so i can power maltbie emg circuit with only one power circuit
So the final pcb work but not as expected so i use solder to be the main contact area but this make the circuit generate a lot of noce in the signal was generated but that wasn't the only problem
The signal wasn't consistent the shape of the generated signal/noise was changing every time I was triggering my Muscles I didn't know why at first but now I can assume that because the surface of the solder wasn't consistent and it had a lot of defects so I'm almost sure that was the cause of this problem
I think in my next version I will use a copper tape as a replacement of electrode placement patch and i think it will be more stable than the solder
In this part, I couldn't do much more at this point due to lack of my resources and of course The deadline
CAD
For the cad, I tried to draw the finger from scratch but my biggest problem is how to make it look kind of realistic so when I coat it with rubber silicone it locks realistic so I modified an STL finger to be easier and faster to implement the mechanism
It worked but not as expected, motor torque was weak and it was so slow but at the same time I was testing with 9 votes battery so maybe it didn't have this match of juice, so I tested it after a while with a decent power supply but the same result the motor was very slow and even there was no consistency in the motor motion, and after a lot of trials, the gear part has been damaged and also there was a little be of fixable problems in the cad model
The next problem was how I'm going to know the position of the motor. I found multiple solutions but the whole Idea became useless after the gear dc motor failed my expectations.
For the New version of the cad model, I will use a servo motor instead of a DC motor and I will try to fix most of the part problems I have faced in the previous design
but I had multiple problems with this design first the servo motor file that I downloaded from the Grabcad website to creat my modifications based on it had a problem the servo design was a little bit small and that ended that the servo motor I bought didn't fit and got damaged in the prosses
The next problem something wrong did happen with the fusion 360 the last version of the design somehow it didn't save probably I can access all the past versions so I don't know the cause of this problem but all I know that I have only the 3d printing files for this version
networking and communications
for this part, I designed a PCB to communicate with the servo motor based on the EMG signal so because of the EMG signal value between 0 and 5 voltage.
I can use a simple command like "map()" to Re-maps the value of the EMG sensor to would get mapped to Servo angle the code would be something like that.
myServo.write(map(analogRead(EMGPin),0,1023,0,180));That what I was going to do if everything worked as expected/0^0\