Research: Dual-Band Transcutaneous Energy Transfer
My undergraduate engineering clinic research is focused on wirelessly delivering power and data between the outside world and implanted medical devices. By the end of my senior year, I hope to have a working prototype for a target application. The most important issue with transcutaneous energy transfer (TET) is transfer efficiency. Inefficiencies in energy transfer manifest in the skin as heat, and raising the temperature of skin more than a few degrees can be dangerous. Therefore, it is vitally important that the wireless power transfer (WPT) is designed to be as efficient as possible, let alone good engineering practice.
From a top-down perspective, we know that the efficiency of energy transmission between devices must be as high as possible. With this, we can infer that the coils must have a very high coupling coefficient. In order to predict the coupling coefficient of a pair of coils, we need to be able to predict the quality factor of a single inductor. This lead me to the first step of my research.
Single Inductor Model
It quickly became apparent in my preliminary research that a flat spiral coil would be the likely best choice of inductor geometry. I began building a MATLAB model to predict the best coil geometries that would fit inside reasonable space constraints. The model was heavily based on the Harold Wheeler approximations in his "Simple Inductance Fomrulas for Radio Coils" (seen right) as well as some more modern approaches to predicting flat and multilayer spiral coils. It predicts the inductance, series resistance, quality factor, and required length of wire for any wire thickness, spacing, inner diameter, and number of turns. Once the model was sufficiently refined, it was used to select a few geometries which could characterize its accuracy.
MATLAB Model of Flat Spiral Inductor
Winding and Measuring Coils
The select coil geometries then needed to be wound and tested in order to check the validity of the model. In order to do this, a coil jig was designed and machined to produce coils with a 15mm inner diameter. Although the inner diameter was fixed, the wire thickness and number of turns in the coil could be varied to change its quality factor. A pair of identical coils was wound for three different geometries. The coils were measured using an LCR meter and the results were plugged back into MATLAB for comparison with the model.
Modeling the Coupling Efficiency Between Two Coils
The single inductor model could now become the foundation of a couplng coefficient model to predict WPT efficiency. After lengthy research, the MATLAB model was expanded based on both long standing and modern techniques. It now predicts within certain geometric constraints the effect on WPT of relative wire thickness, number of turns, inner diameter, and separation distance between two coils. In the future, this model will be used to select the most efficient pair of coils for providing wireless power to implanted medical devices.
MATLAB model and resulting wireless power transfer prediction
Developing a 3D-Printed Test Apparatus
The next step in my research was to develop a testing stand in order to produce experimental results. The test apparatus at least needed the ability to change the geometry of the primary and secondary coils, but looking ahead I designed it with the ability to simulate and measure the effect of different imperfect spacial conditions. Specifically, the test apparatus allows for the adjustment of separation distance, planar misalignment, and rotational misalignment. Additionally, I developed a cartidge system which allows coils to be very easily installed and removed into the stand. The design was created in SolidWorks (right) and printed with ABS filament. A pair of small interfacing circuits was designed to allow the stand to interface with a spectrum/network analyzer, function generator, oscilloscope, and other test equipment. The circuits also support easy swapping of resonant capacitors and load resistances.
Measurements and Future Work
The test stand was used to take measurements which showed promising results and intuitive trends. After some more measurements and testing, the next step in this project will be to incorporate a second pair of coils into the system which will operate on a different band of frequencies (hence the term "dual-band). One band will be set aside to transmit power to the implanted device, while the other will be dedicated to passing data back and forth. More research and modeling will be done to determine good methods to avoid and compensate for cross-coupling between channels while maintaining excellent WPT effeciency. I am very excited for the opportunity to work on this project and I hope you enjoyed reading about it!
