Build, Test, Document
Table of Contents
Final Documentation For Next TeamPort Documentation
The above are links to the final documents from this project, for use of future teams to improve on our design. Future improvements are suggested in these documents.
Electrical Prototype, Final Design, Testing
PCBs were made using PCB Artist Software available for free download from their website.
PCB Prototype DesignThe prototype circuit is shown below for the internal and external boards. After realizing some mistakes, changes were made from this design. Specifically, the step up converter was changed after designing this, more diodes were placed to drain excess charge during switchover, relays were changed, and the internal 5 V regulator was not needed because it can be found on the motor controller.
PCB Final DesignBelow is the final PCB design for the system. All of the components have been moved closer together to take up less space and needless parts have been removed from the prototype. The LED display has been positioned to fit in the external case for the user to see. The layout is also shown below.
Battery Testing ResultsBattery testing was performed during week 6 of MSDII. As the lifetime was expected to be between one and three hours, group members took turns watching the battery and recording the time the battery could stay on and how long the battery took to charge. The results are available in the following Excel file. It was determined that the batteries could last roughly two hours and the charge time was 56 minutes. Based on this knowledge, small batteries were ordered (AA size) to use instead to conserve space and still meet the required spec for lifetime. Battery lifetime testing with these will be done to ensure that they can last long enough.
Also, using the original life time and voltage data, the voltage and time remaining can be plotted for use in the microcontroller. Plotting that data gives the picture below and adding two best fit straight lines to the plot allows the approximation of the time remaining with equations. These were then used to make a look-up table in the code to display the remaining time to the user. There is also a 15 minute safety factor built into the equations, meaning that when the display says 0 minutes left, there is actually almost 15 minutes left, just as a precaution.
Microcontroller ProgrammingExternal Microcontroller Program Data
The internal microcontroller’s primary purpose is to send a periodic signal to the motor controller in the event of a disconnect in order to run the pump. This is performed using the Timer A module available on the microcontroller. In this setup, Timer A is run in continuous mode and counts all the way to 0xFFFF. The register TACCR0 is used to determine the width of the actual pulse and the output is 3.3 V during the duration of 0 to TACCR0. From TACCR0 to 0xFFFF the pulse is 0. With this method, the correct duty cycle can be achieved needed by the motor controller.
This microcontroller is also in charge of monitoring the internal battery during disconnect. This is done by monitoring the voltage of the battery using the analog to digital converter on the microcontroller. This information is then put in an ASCII string and sent wirelessly to the external microcontroller.
The final function is to monitor the temperature of the circuit during the charge cycle, as this produces the most heat. This is performed using voltage division between a 10 kΩ resistor and a thermistor rated for 10 kΩ getting power from the 5 V regulator such that the output is nominally 2.5 V. This voltage fluctuates based on the resistance of the thermistor which changes with temperature. This voltage is also read using an analog to digital converter and compared to an equation to determine the temperature.
The primary purpose of the external microcontroller is to receive the battery voltage data using the UART and the receiver and use the voltage to determine an approximate lifetime remaining. The lifetime (in minutes) is then displayed on the two 7-segment displays for the user. When there are 30 minutes remaining, one beep from the alarm will occur. When there are 15 minutes, two beeps will be given. When the lifetime reaches zero, a solid beep is heard until reconnected. There is also a 15 minute safety factor built in, such that when the display says that there is zero minutes remaining there is actually 15 minutes of runtime left. This microcontroller also has access to a relay, which is used to toggle the alarm on and off as needed.
Mechanical Testing and Results
Magnetic Force Test DataMagnetic Force Data
The data from magnetic force testing is given in the above Excel sheet. The test device used was constructed from mechanical components and is shown in the picture below using a force gauge.
Heat AnalysisHeat Analysis Report
Above is the final heat analysis for the internal device implant. Full details of the analysis are in the document. The conclusion is stated below with reference to figures in the document.
Initially, the energy equation was applied to the surfaces of the device. Figure 1 shows the Heat Rate for each surface. One data plot shows the dissipation to the body and the other shows the heat dissipation to air. Adding these values reveals the total heat dissipation from the device for various inputs of heat generation. The lines represent the total curve while the dots reveal the region that is within the temperature range of the device that will not cause harm to the body. The maximum heat generation allowed from this analysis appears to be about 0.6W.
To follow the worst case scenario, one would look at the plot of heat generation with respect to the temperature of the air. This shows how much heat will be dissipated when we assume that there is no heat dissipation into the body (adiabatic). The maximum heat generation allowed from this analysis appears to be about 0.35W.
The relationship between temperature and heat generation (W) for a range of 0 to 0.5 Watts of heat generation
The heat equation was solved to determine the temperature distribution from the device to the clothing surface for multiple heat generation values. Figure 2 shows the relationship between temperature and heat generation (W) for a range of 0 to 0.5 Watts of heat generation. We can see from the figures that the temperature of the device facing the inside of the body remains near 40 degrees Celsius when the heat generation is below around 0.4 Watts. Table 1 shows the heat flux values for up to 2 Watts of heat generation. This data shows that the device would safely be able to generate amount of heat energy while remaining below our constraint of 40mW/cm2.
In general the temperature distribution solution seems to fit a model that we would expect better than the previous solutions. The problem still remains that on the right side of the device, the initial temperature and heat generation of the human body is not accounted for. Also, this is not entirely the worst case scenario because if the left side of the device were to be adiabatic, then the temperature would be unknown and thus, a solution could not be obtained. Also, this model neglects the heat generation in the body but assumes that the body’s energy was used to bring the device temperature up to body temperature initially where it then reaches steady state and can then be neglected in the solution.
Both of the solutions results in relatively high temperature values for low heat flux values. As we are not satisfied with the results, a solution will be pursued in ANSYS. Figure 3 is a contour plot of the heat flux. If the values are reduced to mW/cm2 then the result would be a tenth of the values that are displayed. This means that for 2 Watts of heat generation, the maximum heat flux from the device would be 16.6mW/cm2. Looking more carefully at the plot we see that this occurs only at the corners of the device. In the body the maximum heat flux is actually only about 11.2mW/cm2. The 2D analysis results show that a much smaller heat flux can be obtained than from the 1D analysis for the same amounts of heat generation.
In Figure 4, the temperature appears to rise relatively high. This is not likely to pose too much of an issue when we look at some of the assumptions that we’ve made. We assumed that the body was at steady state and that the device was only an additional heat source. Thus, the analysis doesn’t include the metabolic heat generation of the body nor does it include the perfusion rate or circulation of heat due to the blood. In theory, this analysis seems quite simple but there is certainly a lot going on inside the body that is truly difficult to model. The result of this analysis has many flaws and is likely to have a large amount of error. However, it is a good model to be used in order to get a grasp of what is happening.
From the data, particularly the ANSYS model as it is likely to be more accurate since it requires fewer assumptions, we can conclude that the device is not likely to pose a threat to the body when generating up to 2 Watts of heat energy. The factor of safety between the results and our limitations is about 4 which leaves an allowance of more heat generation within the internal device.
Heat TestingTest results from the heat analysis are shown in the picture below. It was determined that the heat flux was 46.2 mW/cm2. With longer run time, this value would increase. Either way, the heat flux is out of our desired operating range of 40 mW/cm2.
Internal CasingThe internal casing design is shown below.
Required Parts - BOM
The above is a link to all of the parts ordered during testing and implementation of the internal and external boards. After the prototype was made, some of the components were changed, added, or removed. The final list of electrical components and their reference number on the final schematics is given on the second page of the Excel document. The final cost to completely build one of these devices is roughly $250.
Video Demo With P13021 and Connection AgreementBelow is a video of our internal product functional with P13021 and out operating agreement with them.
"Right Click --> Save As" to view. Then open on PC.