Project
Replace older systems with more sophisticated sensor systems.
Old system
The biology department at NAU has an existing sensor system to monitor and control the temperature and pressure of gas chemistry reactions for radiocarbon analysis. This embedded system contains a set of 12 gas reactors that each have a small pressure sensor to monitor the reaction.
The Problem
The problem is that the model of pressure sensor we use is no longer made, and the newer models have a different input voltage and return voltage, a different number of pins, different spacing of pins, etc. They'd like to start using the new sensors, but the interface was not made to work with the newer sensors. They would also like to integrate the heating system and the pressure sensor system, so that the heaters can be automatically switched off after the pressure has dropped.
Our Solution
The initial prototype version of the gas reactor system utilizes an Arduino Mega and twelve AMS5812 absolute pressure sensors. The team has chosen to work with an Arduino Mega due to the familiarity of the programming language and low current requirement of the system. Another advantage of the Arduino Mega is a high number of analog inputs that can allow for a twelve sensor array. The AMS5812 pressure sensors were chosen by the team sponsor before the beginning of the project because it was recommended and preferred by other chemists that need to monitor pressure. These sensors have the capability to read from 0 to 30 PSI (pounds per square inch).
Project Introduction
We appreciate being selected to work on the Gas Reactor Sensing system, under the sponsorship of Christopher Ebert. DeltaP worked effectively and efficiently to develop an upgraded pressure monitoring system for the Northern Arizona University science department. Although we were negatively affected by global circumstances beyond our control, we have ended our capstone program with a functioning and mostly complete final product. In lieu of the parts necessary to complete our design, we have provided thorough documentation regarding which features need to be finalized by another capstone team, or our sponsor. For our capstone project we worked in tandem with the NAU environmental science department to construct an automated gas sensor reacting system to measure deposits of Carbon 14 within atmospheric samples. We were lacking a full understanding of the data research being collected through the use of our embedded system, but we found it crucial to complete our project to the highest standards asked of our team by our project sponsor, Christopher Ebert. Dr. Ebert is a PHD researcher with a background in chemistry and his work involves the study of the displacement of Carbon 14 particles distributed throughout many regions of the globe, including Coconino County. The gas reactor system is a unification of multiple disciplines within the science community to help better understand the anthropological effects of humans upon the natural environment. Researchers within the fields of environmental sciences and chemistry have been developing various methods for extracting Carbon 14 from atmospheric samples. Samples of Carbon 14 are collected in sealed-glass pipettes that are cracked open within a sensor unit attached to a vacuum chamber. The Carbon 14 is extracted from the atmospheric sample held in the vacuum chamber. The internal pressure of the system is brought to zero PSI (pounds per square inch) causing the Carbon 14 to collect at the bottom of the glass pipette. A small graphite crystal is positioned at the bottom of the glass chamber, and under vacuum pressure the Carbon 14 collects around the graphite. Following the extraction process, a heating unit solidifies the carbon onto a graphite crystal. A wide range of data can be interpolated from Carbon 14 extraction, a primary use of this practice is gauging carbon emissions over time. Our capstone project is related to a gas reaction process that involves vacuum chambers and heating units. For our project we will not be involved in the testing equipment related to the graphite crystal that is formed by the reaction process. The basis of our capstone project is to work with a carbon pressure system in order to remove carbon from the air via a vacuum pump system, and test the emissions and air quality. With this in mind, we are specifically working with the sensors. However, the general knowledge of the entire system is vital for our understanding.
Project Requirements
The gas sensing reactor system must accurately read pressure levels within a vacuum and output the results to some form of monitoring system. As the development of the project concluded, other design features that had been requested by the project sponsor were no longer feasible due to shipping limitations. These additional features would have been a heating system with wireless communication for carbon conversion as the final stage of the gas reaction process. Other features that have been implemented are a solid aluminum housing and modular sensor inputs.
Poster
Link to Capstone Video
https://www.youtube.com/watch?v=EQKh59YbbVU
Subsystem
Pressure Sensing System
The system is controlled by an Arduino Mega and utilizes an array of thirteen AMS5812 pressure sensors. Our team initiated the process of fabricating the final system for the pressure monitoring system by designing and milling our own printed circuit boards. Using a CNC router, we have developed circuit boards for the integration of physical controls and wireless systems. KiCad was utilized to design the boards. The system has an LCD display and physical controls on the face of the device. The embedded system is housed in a modestly sized aluminum case. A spreadsheet had been implemented to keep track of our project budget in real-time. Delivery of parts had been slow and a shipment of parts was lost before it could be received by the team. As a result, final assembly components were integrated into the project prototype as parts arrived. To make our deadline, we had to revise several design aspects. Functionality of our circuit boards are described in further detail in the system manual. The pressure sensing system is broken down into subsystems below, listed with which team member headed each task
Wireless Communication System
For the wireless communication system our team has chosen to use a modular add on, or shield, called Zigbee. This part will be added onto our microprocessor, the Arduino Mega, and allow it to communicate wirelessly. The teammate that is in charge of this section of the project is Mateo Zhang. At this moment in time our team was able to get the device functioning with our final design. We were able to send the data of various pressure sensors being used by our device to another device that had a Zigbee USB device connected to it. This will allow our client Chris Ebert the ability to monitor the state of the pressure sensing device remotely from his office. Since the time it takes for these Carbon reactions can vary from two to four hours our client expressed how helpful it will be to not have to continuously keep going to the lab and checking on the system. Since our team has been able to send and receive signals from our prototype design, we are currently looking into a way for the data being sent to be recorded and displayed in a graphical user interface (GUI). The final state of the wireless communication system is functioning but incomplete. Due to COVID-19, Mateo left the United States in the middle of the Spring semester. Any further development of the GUI was halted at this point. A Zigbee wireless module has been incorporated into the final design of the system and is coded to transmit data from the thirteen pressure sensors. It will take time to make a GUI that satisfies the requirements of the sponsor. Remaining team members used the remainder of our time to complete the pressure sensing system. At the technical level of ZigBee, its code is also written on Arduino. I use xctu software to set up two xbees. Two xbees are set on the xctu, one is regarded as the receiving end and the other as the sending end. Set the address of XBee on xctu. In terms of hardware. We connect XBee to the computer and Arduino through purchased accessories.
Pressure Sensing Array
FAt the request of our project sponsor, we utilized an array of thirteen AMS5812 absolute pressure sensors. These sensors are highly accurate with an input voltage of 5 volts. We chose the AMS5812 model A sensors that can detect a pressure range between zero and 30 pressure per square inch (psi). Stipulations of our project require the ability to detect vacuum pressure (0 psi), relative atmospheric pressure (about 11.7 to 12 psi in Flagstaff) and up to 25 psi. The system uses twelve sensors for testing samples and a thirteenth sensor to use as a reference on the vacuum seal. We chose to use the internal analog to digital converters of the Arduino to read the sensor values because the I2C ports of the Arduino are being utilized by the Zigbee wireless communication unit. Initially, there were issues regarding the accuracy of our sensor readings. After refining our code we were able to get the accuracy within 1000th of a percent. The large sensor array cannot be powered directly from the Arduino due to the embedded system over-drawing current. An important aspect of our system that we were not able to implement due to restrictions on ordering parts is a voltage regulator. The sensors should be separated into groups of six and seven sensors, with each set being provided an input voltage through independent voltage regulators. The system without voltage regulation only has enough current to power either the wireless communication unit or the pressure sensors, but cannot power both simultaneously.
Housing and Controls
Our printed circuit boards were developed in house using software and tools provided by the NAU engineering department. There was quite a bit to learn regarding PCB design, such as trace width tolerances, program literacy, milling speeds, and drilling depth. After working with multiple circuit design programs, we ultimately ended up using KiCad to develop our boards. The boards were then milled using a Bantam CNC router. After printing the boards, the team worked together to solder components and assemble ribbon cables. Two boards were designed, one for the physical controls and another used as a link between the Arduino and peripherals.
Heating Control Unit
For the heating control unit our team has figured out a way to control the heating units provided to us from our client. However, the difficulty in this section of the capstone project stems from the potential risks that can come up depending on how this part is developed. For this reason the team has been carefully debating the different routes we could take in order to keep this part of the project as safe as possible. One route our team was debating was to have a separate heating control unit for each heating unit, this would mean that we would need 12 to 13 heating control units. Looking at our team’s budget and the cost of heating control units this idea was not possible in terms of cost, even though we liked this idea the most since it meant that each system would be isolated from each other which is good given the high current that would be running through each heater. The other plan we thought of was to have a relay system that could switch to different heaters and heat them up. The danger of this though is it could be difficult to keep track of which heater is currently live and which ones aren’t. In the end we talked to our client Chris Ebert and decided that the best course of action would be to get a heating control unit working so he would at least be able to use one. This would mean that this section of the capstone project would likely go to some future capstone team. That’s currently where our team is at in regards to the heating control unit subsystem of the capstone project.
Project Summary
Initial testing and prototyping had been completed during the Fall semester. During the Spring, each member of the team was independently focused on core aspects of the project. Gabriel worked on PCB design and housing, Cameron worked on the system code and optimization of the physical controls, Mateo worked on the wireless communication and web development, and Ben worked on the peripheral heating array. Unfortunately, midway through the semester our project was negatively impacted by the spread of COVID-19. This slowed down our design development and completely halted our ability to order parts. We completed the pressure sensing unit with parts on hand but had to abandon any efforts towards developing the heating array. The limited use of campus facilities during the COVID-19 pandemic has drastically slowed down production, becoming the hardest challenge standing between our team completing the project. Our team came in under budget at the end of the Spring semester. Due to social distancing guidelines our ability to order parts was halted. After researching the cost of parts, we knew that the heating array was going to be the most expensive aspect of our project to produce. Since we did not develop the heating array, we finish with about half of our budget left over. We also dealt with the unfortunate issue of a lost shipment, resulting in an unfinished array of pressure sensors. The gas reactor sensing capstone tied together many aspects of our undergraduate engineering program. The development of an embedded system required coding, controls, wireless communication, and circuit analysis. Our team worked effectively towards completing our task and are able to provide our sponsor with a mostly finished product.
Our Team
Gabriel Bixler
Team Leader, Sponsor Liaison and Scheduling CoordinatorC, C++, Matlab, Assembly, Arduino, Circuit testing, troubleshooting, Soldering, Technician experience, Multisim, SPICE, Solidworks, Experience with ecology.
Cameron Divine
Treasurer, Faculty Technical Advisor Liaison and Vendor LiaisonC Programming, Python Programming, Circuit Testing, Circuit Simulation, Solidworks/Autocad, Soldering, Cryptography.
Benjamin Shugrue
Document Coordinator and and Presentation CoordinatorC, MATLAB, Circuit Testing, Can read electrical and pneumatic schematics, Soldering, Vacuum Experience, Wiring knowledge and ability.
Qian Zhang
Secretary and Project web site coordinatorC, java, Matlab, Multisim, Soldering, MSP 430, Electromagnetics, Date structure.
Mentor
Christopher Ebert
The Ted Schuur lab, in the biology department, studies permafrost carbon in Alaska.
More about Chris Ebert
We manage a permafrost warming experiment and observe landscape degradation, carbon loss in soil, and plant response to warming. We are buying an Accelerator Mass Spectrometer (AMS), a cutting edge tool for analyzing radiocarbon (e.g., carbon dating). The AMS will be analyzing our samples and samples for any NAU user, and potentially outside clients if we have spare capacity. We want to improve our analytical tools and sample preparation techniques as this lab takes off – we plan to complete with the most powerful AMS labs in the world at lower cost by using simpler operations and more automated techniques. Chris Ebert is the manager of the Ted Schuur lab. He runs the sample preparation laboratory where the gas reaction system operates, and he is excited to start running the AMS after installation in spring 2020. This new sensor system will process about 3,000 samples per year with better performance than our old system, improving our science and helping us attract users.