With the start of my senior year at Western Washington University in the electrical engineering department, comes the senior project. For my senior project, I have decided to make a concussion sensing module for use in sports. This idea came to me after thinking about my own personal experience. While playing football my junior year in high school, I suffered a hematoma. I did not lose consciousness, but instead my coaches saw me randomly laughing while playing defense and pulled me out of the game. After sending my cat scans to Harborview Medical Center, they found a area of a blood which then required me to be airlifted to Harborview. Luckily after arriving and doing another cat scan, the area seemed to be clear. Looking back at the film after, we determined the hit that caused the injury. It was about eight plays before I was taken out of the game, during an offensive drive. As quarterback, none of my teammates saw any change my game play. I wanted to come up with a device the continuously monitors athlete’s head movements and alert coaches or officials of possibly brain injuries. I also wanted to be able to capture the data from the device to be further researched off the field.
My goal is to design this device at a lower cost of what is currently on the market. There is still research to be done. The basic functionality of the device right now will be sensing and calculating the change in acceleration and rotational velocity using a accelerometer and gyroscope. If there is a large enough change between samples, an alert will transmit a signal to a external device (iPad, computer, etc). The device will resemble a running headphone. It will have the a rubber piece that will attach the earpiece and have a in ear piece that will be used for rigidity.
Before getting to any of the design, I needed to selected components that would meet the goals of the project. I started out by picking out a microcontroller (MCU) could that meet the requirements of the project. Initially, the NXP KW41Z was the MCU of choice since it considering most of the peripherals would be using SPI or I2C and had plenty of ROM for storage of data. After doing some further research, I found the Rigado R41Z. The Rigado R41Z used the NXP KW41Z and also enabled Thread with IEEE 802.15.3 and Bluetooth Low Energy connectivity. The R41Z also has a antenna trace on chip, which was a large perk considered I did not have access to a network analyzer. At the time of purchase, the R41Z cost $15.10 and is the most expensive part on my board.
With the MCU selected, I could now search for the rest of the components. These components need to be able to be supported by the on-chip resources on the MCU. I started out by looking for an accelerometer. The accelerometer and the gyroscope would need to each of 3 degrees of freedom to cover any possible movements. I also consulted my professor on some important specifications to look for specific to my project and bandwidth was another large importance considering head movements could happen in small time frame. I went back to looking at components and found that accelerometers, as well as gyroscopes with higher bandwidth, are quite a bit more expensive than accelerometer with other similar specs. I then came across the Analog Devices ADXL375, which had a bandwidth of up to 1 kHz and allowed for high resolution of measurements up to +/- 200 g. Though +/- 200 g might have been a little overkill, the bandwidth and the shock threshold abilities of the device sold me. The only problem with the device was the price. The ADXL375 at the time of purchase was at $10.95, which would come out to be the second most expensive component, behind the R41Z.
Shortly after I found the ADXL375, I started to look for a gyroscope with comparable bandwidth since my plan is to have the same output from each device for easy comparison between data. I came across the InvenSense ITG-3701 which had a data sample output of up to 32kHz, could measure up to +/- 4000°/sec, and a 16-bit ADCs for digitizing sensor outputs. At the time of purchase, the cost of the ITG-3701 was $5.58.
Realizing that there was going to be six streams of data being stored, I did a quick simulation of data in MATLAB and found that internal ROM on the R41Z was not going to be enough and external flash would be needed.
2D Top Layer
2D Bottom Layer
3D Top View
3D Bottom View
3D Angled View
After getting a review from my professor, it was time to get the gerber files off to fabrication facility. We used PCBWay and it was a very quick turn around and was very happy with the results.
Finally, all the components that have been sitting in my locker were about to make their way onto the board. I did not have a stencil made by the fabrication facility, but instead used our department’s stencil cutter. Using the top paste layer gerber file and a modified Cameo cutter, I ended with a nice stencil shown below.
After a flux wipe down, the board was ready for solder paste. To keep the board and stencil steady, scrap PCBs were used to hold the two in place. There was a little technique needed to get the solder paste to spread correctly.
With the size of my board, along with the density of components on the board, I was glad that my department had a pick and place machine.