Michael Spencer

SAMPLE 1: Ion Stability Calculator

This app started out as a C++ Windows Application. As we started getting more systems built I wanted to make something more portable to the different computers. This also made it easier to change on the fly for both bugs and features. Doing this version gave me a chance to learn HTML 5. It is based off of Mathieu Equations. In this case the Mathieu Equation is determining the stability of an orbit of an ion. In the case of a 2D ion trap there is stability in the cylindrical z direction and stability in the radial direction. The red lines left to right represent a beta (B) from 0 to 1 for the z motion. Anything below 0 and above 1 would represent an unstable orbit, however 0 and 1 themselves are barley stable . The same goes for the blue lines except for the fact that it represents the r direction. The x axis is a q value that is determined from the mass of the ion, RF electrical characteristics, and geometry of the trap. The y axis is an a value that is determined from the same except using DC electrical characteristics instead of the RF.

SAMPLE 2: Ion Simulator

This is a very stripped down version of the windows application. It was designed solely to be an example of what I did in the full scale application. The windows version has a few more views to give various different information on the performance of the trap (Spectrum, Ion Ejection Location, Electric signal, etc.). This web app utilizes WebGL for the bases of the simulation. It takes it a few seconds to load the files that it uses for determining the fields. The simulator was created in order to save time on simulating various different geometries and fields applied to the trap. A commercial product SIMION is typically used for this. The problem we had with it was the timing of each simulation was extremely slow. For an initial time estimate it was about 36 hours using SIMION and was around 10-15 minutes for ours. To be completely fair, SIMION was single threaded and mine was multi threaded (thanks to the software team for setting this up for me). To make it fair we opened 6 instances of SIMION and spread the ions between the 6 instances of SIMION and got the same simulation down to about 6 hours (still too long). Again using this application does take some know how of ion traps, but the web app will get the point across.

SAMPLE 3: CAD (Coming Soon)

I would love to show past models, assemblies and drawings that I have worked on for the past 9 years, but those are company proprietary.  I will have to resort to making something up for myself.  At 1st Detect we had to deal with some very tight tolerances.  The problem with shrinking the trap is that any imperfection in the surfaces caused field differences that the ions saw.  This would cause an unsymmetrical field and have higher order resonant points.  For most parts standard tolerencing was sufficient, but on the trap we also maintained fairly tight GD&T in order to ensure that all the traps were as similar as we could get them.