Understanding Modeling Software

(3/14/17)

We have added to our understanding of resonant cavity design by configuring and executing Superfish 2D models.  This has allowed us to quickly explore transverse electromagnetic (TE) modes present in the Cavity in order to interpret the actual two-port parametric measurements we will make using a vector network analyzer.

Preforming Measurements On Unknown Sample

(3/29/17)

After receiving the resonant cavity from our sponsors, in addition to a unknown sample to test, we began playing around with the measurement process. After getting familiar with the Vector Network Analyzer made available to us by our advisor, we were able to look at the frequency response of the material. Our results were repeatable and we were getting a similar frequency response to that of our sponsors.

Determining the Quality Factor and Dielectric Constant of Unknown Sample

(4/10/17)

Once we were getting repeatable results, we were then able to extract important parameters from the Vector Network Analyzer (VNA) measurements: quality factor and resonant frequencies. We then used our 2D model on Superfish of the cavity, to determine the quality factor, with correction, and the dielectric constant. The correction factor for the quality factor is necessary, because what we get in our measurements includes the cavity itself, so we use the correction factor to obtain the quality factor for just the sample. The dielectric constant is determined by playing around with the dielectric constant for the sample in Superfish, until we get the response on Superfish to match what we found in our measurements.

Realizing the Improved Design

(4/15/17)

We were able to realize our improved design for the cavity by trying out different dimensions for our cavity on the software, keeping the sample size and orientation the same. Our sponsors gave us a target number of modes in a specified frequency range, and we changed variables on Superfish until we achieved this. Once the target number and modes and range was reached, we were then able to verify that the efficiency improved. This was also done on Superfish by looking at the ratio of energy in the sample over energy in the cavity for our various modes. 

Submitting the Improved Design to be Machined

(Summer 2017)

During the summer, we finalized the design in Solidworks so that we could send it to be machined before MSU fall semester began. From there, we could focus on accomplishing some of our other goals while we waited for the cavity to be ready for testing.

Realizing the Automation Software

(9/1/17)

Due to the addition of automation in the level one requirements, we needed to quickly decide what software would be optimal in instrument control. With the guidance of our sponsor and our advisor, our team decided that LabVIEW would provide a nice GUI for the user to run measurements. The choice was also influenced by the accessibility of instrument drivers available on LabVIEW that were not all available with Matlab. We were still flexible on the software choice if we wanted to process the data using a combination of LabVIEW and Matlab.

Progress in COMSOL Model

(9/15/17)

Spending the end of the summer understanding COMSOL software, Jessie started making progress on the COMSOL model for the improved cavity design. She did so by comparing our 3D model on COMSOL to the 2D model  on Superfish.

Progress in Automation Code

(9/30/17)

While waiting for the cavity and LCR meter to arrive, Shane and Chris worked with the instruments that we had available: Dr. Becker's VNA and Watlow environment chamber controller sent by our sponsor. We were able to make progress in writing the instrument code for the VNA that could be tested on the cavity upon arrival. 

Validation of Design

(10/15/17)

Prior to sending the cavity to MSU, our sponsor wanted to know what we should expect to be seeing so that they could help us understand the measurements. Jessie aided our sponsor with taking measurements on the revised cavity. At the same time, Jessie continued to work on our COMSOL model to verify consistency with COMSOL and Superfish. 

Further Progress with Automation Code

(10/15/17)

Once validation was complete, our sponsor sent us the cavity, the LCR Meter, and samples to run tests on. To begin with, Shane started to work on the LCR Meter code in order to integrate the instrumentation code into one virtual interface on LabVIEW. At this point, there was still difficulty in initializing communication with the environmental chamber controller on LabVIEW.

Starting to Process Data

(10/20/17)

Quickly, the team realized how much raw data our sponsor would have to go through if we did not find an efficient way to process the data. From the raw data, our sponsor already had, Jessie wrote a Matlab script that would graph appropriately and find modes. We wanted to adapt the desired output from Jessie's script into a more automated process that was taking in data the way we were saving it via LabVIEW. Also, we did not want to manually find peaks, we wanted the code to do so for us. We began to test various Matlab functions to properly detect the peaks.

Validation of LabVIEW Automation Code and Matlab Peak Detection

(11/1/17)

After ensuring the proper orientation of the samples in the LCR Meter and VNA, we ran a series of measurements by using the LabVIEW code to acquire data on the samples. With the LCR Meter measurements, we were able to validate that LabVIEW was collecting and saving the correct data by observing the desired values on the instrument front panel. For the VNA, we would have difficulty in inspection with that method; so, we acquired the frequency traces manually off the VNA. From the manual and LabVIEW traces, we plotted both to inspect discrepencies. 

Since we needed a way to process data in this validation step, the team found this as an opportune time to test the peak detection code we were working on. 

Integration of Full System Testing

(12/5/17)

Prior to presenting the design at Senior Design Fair, the team completed full system testing of measurements. Through the use of our Labview Program, Shane was able to run automated tests through a simulated range of temperatures for the samples in the cavity, designed by Jessie, and in the LCR Meter. From Shane's Matlab script for peak detection, we were then able to utilize Jessie's excel file that converted measured values into corrected values for the sample. We found that there was less than 10% experimental error between our sponsor's data and the data collected by Shane.