Since mid January I have been conceptualising an engineering oriented valentines day gift for my girlfriend. This was a simple yet challenging product to conceptualise and design which had a deadline, partially simulating real world projects. This was a collaborative project with my friend Chris who has a strength in 3D modelling and design. 

My contribution consisted of programming the ESP32 to control a servo and LED sequence in response to button presses. When the button is pressed, the servo rotates 90 degrees, triggering a pulsing LED effect. Pressing the button again stops the LED sequence and returns the servo to its original position. My previous project enlightened me to the world of PWM signals which is how the servo and pulsing of the LEDs function. However, I encountered a challenge during programming - the microcontroller couldn't generate two different PWM signals simultaneously. This created a significant hurdle in getting the servo to rotate while the LEDs pulsed at the same time. After hours of failed attempts, I finally encountered a very simple solution. By using the detach function in the Arduino IDE, I could detach and reattach the servo when needed, allowing the ESP to produce the appropriate PWM signals at independent intervals. 

Chris designed and 3D-printed the heart-shaped box shown in the image carousel. The model consisted of three main components: the lid, the outer casing, and the inner casing, which housed the electronics. In the rush to complete the project, we overlooked adding proper openings for the servo and button, and the hinge mechanism had a hole too small for the intended metal rod. As a quick fix, we melted holes and used a 3D pen to fill gaps and secure parts as needed. Chris and I also soldered all the LEDs, resistors, and pins to the ESP32, carefully arranging them within the casing. To open the lid, I implemented a two-point free hinge mechanism using sections of a metal coat hanger.

Unfortunately we forgot to add a charging port, or a method of replacing a battery, which is a major oversight on our part. Thus in following iterations this will be fixed immediately. 

This is a personal summer project over December, January and February 2024/2025. It was initiated to deepen my interest and knowledge into the world of circuitry and 3D modelling which was fostered throughout the previous semester. 

I first started with understanding the Auto Desk fusion 360 software through various tutorials. Extrusion, cutting, sketches and shortcuts were a few of the multitude of skills learnt during this period. My goal of modelling a hinge mechanism was achieved, allowing me to 3D print a functional. rather than aesthetically focused, unit of a finger. After a few more weeks into the holidays, I was more comfortable designing an entire hand based on these hinges. 

When it came time to design a circuit, I had gained an understanding of PWM signals and how transistors, resistors and capacitors are viewed in circuit development. Servos operated on a 50 Hz 1ms - 2ms pulse width signal. This can be produced by either fine tuning a astable multivibrator circuit, or using a 555 timer IC chip and choosing appropriate resistor/capacitor values. Working with analog circuits has proven to be both incredibly enlightening and frustrating at the same time, as the signal it produces is not perfect. As of now, I have not finalized the circuit, with an iterative process required to produce a smooth turning servo without an Arduino. 

A pulse oximeter takes advantage of the fact that hemoglobin and oxyhemoglobin absorb differing wavelengths of light. Thus by applying red and IR light onto a finger, with every change in blood volume corresponding to diastole and systole phases of the heart, 2 absorption profiles can be observed. Combining the peak and trough values of these profiles through a mathematical equation produces an oxygen saturation level for the user. 

This project involved a substantial amount of critical thinking in order to finalize a circuit design that could mimic the operations of a traditional pulse oximeter. A normal resting heart lies around 60 to 80 bpm. Thus a simple passive RC band pass filter was chosen to eliminate frequencies below 1 Hz and above 3 Hz. The voltage and current reading from the phototransistor by itself is incredibly minimal, thus an amplifier must be used. An Op-amp allowed a non-inverting amplification of the input signal, allowing easy visualization on to the oscilloscope. 

I had gained immense experience in oscilloscopes, NI elvis boards and Elvis software. They were used in the testing of every component in the circuit in order to ensure everything was working as intended. For instance inputting a square wave to record the voltage output of the phototransistor, essentially a binary on and off tester. Additionally, experience in LabView was used to extract vital information from the IR and Red LED absorption signals to calculate the oxygen saturation percentage.