Sept '00-Aug '01 — JDS Uniphase Inc., Ottawa, Ontario

  • Product Development in the Optical Frequency Interleavers group of the Fibre Optic Products division
  • Setup an elaborate computer simulation of the primary optical part in our product
  • Worked closely with manufacturing to determine their specifications and tolerances
  • Simulated the product assembly, and determined yields based on customer specifications
  • Qualified the simulation by comparing with actual yields from manufacturing
  • Determined the manufacturability of new designs, and what areas needed improvement
  • Learned about the fibre optic industry through training sessions, courses, and seminars:
    • 2-day course: Optics for Optical Fibre Telecommunications
    • Half-day course: Gaussian Beam Propagation

I was recruited for this job by JDS Uniphase even before I finished my degree. However, I couldn't start until September as I had made a previous commitment to work as a Research Assistant at SFU for the summer. I then moved to Ottawa, Ontario to begin work. I worked there for a year, before down-sizing of the company forced the layoff of many people. The group I worked in was later dissolved, and the entire Ottawa operation was shutdown and moved to San Jose.

I was a Product Development Engineer in the Optical Frequency Interleavers Group of the Fibre Optic Products division. In this job, I provided direct support to the Application Engineers and Product Line Managers for their data collection and analysis needs for this product.

The product that I spent most of my time with was an etalon-based Interleaver, which could split 40 incoming channels in a single fibre optic cable, into two sets of 20 outgoing channels with double the channel separation. Also, the device could be used in reverse to merge two sets of 20 channels into a single set of 40 channels. This device was useful as a component in dense wavelength-division multiplexing (DWDM) systems which transmit large amounts of data over a single fibre optic cable by using a large number of channels.

A spin-off of this product, which I also did a fair amount of work on, was the etalon-based Tunable Dispersion Compensator (TDC). This device used almost the same technology as the Interleaver, except that it passed all the channels, but compensated each channel individually for its dispersion. The amount of dispersion compensated is tunable by an input voltage to the desired amount. For a description of this effect, see these papers:

My area of expertise for these products was in the simulation of their construction to determine attainable specifications for the final device. This first involved the understanding of the manufacturing process and tolerances for the devices and the etalons that are used. This knowledge then had to be turned into a simulation of the construction of a set of random devices. The optics of these devices was then simulated using Gaussian optics and ray-tracing to see how they would perform in a fibre optic system. This enabled me to determine how many devices would pass or fail a given specification, and therefore whether that specification was possible.

All of the programming for this was done in Visual Basic for Excel. This enabled the simulation to also include the automatic analysis of the results and presentation in graph form.

Though most of my work was on these simulations and the resulting data analysis, I did other things as well:

  • experimented with our products in our research lab, using cutting-edge instruments and equipment to test the validity of my simulations on real devices
  • learned to assemble, in a clean room, some of the products that we manufactured
  • attended training sessions in Gaussian Beam Propagation and Optics for Optical Fibre Telecommunications
  • attended the weekly JDSU seminar, in which current and upcoming trends in the fibre optic industry were discussed and debated