It isn't easy defending an IP lawsuit in the Eastern District of Texas.
This was a long-running case alleging patent infringement and trade secret misappropriation. Chipmaking company Intersil Inc. wanted to get into the market for ambient light sensors (ALS). An ALS is a tiny chip that goes into phones and other battery-powered devices, which lets the processor know how brightly lit the device's surroundings are. That lets it adjust display brightness automatically to compensate, which saves battery power in dim surroundings and keeps the display from being too bright or too dim.
The leading ALS chip maker at the time was Texas Advanced Optoelectronics Systems (TAOS), which was a spinoff from Texas Instruments. (Local boys made good: strike one in the EDT.) Intersil approached TAOS about an acquisition, and was favourably received. The two sides did their due diligence, and Intersil made an offer to purchase TAOS. The offer was too low for TAOS's taste, owing to Intersil's having found out that TAOS's biggest customer was going to pull the plug soon, so that there was likely to be a serious revenue hole. TAOS asked for a higher price. Intersil sharpened their pencils and crunched the numbers a different way, but the resulting number was still too low for TAOS's liking, so the deal fell through and TAOS found another way to get the money to ride out the revenue shortage.
Awhile later, Intersil came out with its own line of ambient light sensors. There were some similarities between them and TAOS's devices, so TAOS sued, asking damages in the $100M range. As Intersil's testifying expert on the trade secret side, I was able to show in my reports, deposition, and court testimony that Intersil had developed its devices entirely independently, based on its earlier high-speed integrated photodiode work, and had not derived any unfair advantage from TAOS's technology. TAOS's account of its alleged secets was iunaccountably vague throughout. Although both I and the legal team were pretty confident of a good outcome, the jury found Intersil liable for $56M, later raised to $77M on appeal.
Reconstructing, it seems that the jury never got past the fact that (due to a succession of apparently minor screwups) Intersil had retained TAOS information that they weren't entitled to. Some was in the untouched-but-undeleted email account of a former business development manager--the one who signed the nondisclosure agreement with TAOS--and some was in the archives of the Intersil legal department. It seems that this manager had used a different contract from Intersil's standard NDA but hadn't run it past the legal department.
Intersil's usual NDA allowed them to keep a copy of the confidential information archived in their legal department for use in case of a dispute, which is a common kind of provision in NDAs. However, this contract obliged Intersil to destroy or return all confidential material at the termination of the agreement. Intersil legal didn't know this, so they kept a copy as usual. Although their engineers didn't use them for anything, those unfortunate stray copies were very expensive for Intersil. A pity.
Drift and 1/f Noise at 70-90 °C
Sometimes you have to find out things that aren't in the datasheet, and even the manufacturer may not know.
Thermoacoustic fridges are magic: you heat one end, and the other end gets cold. (Of course you have to sink all that heat from the middle.) They can easily be made long and skinny, and so are a natural for use down drillholes. They're also made entirely of metal, and have no moving parts, so they will survive bouncing around in the back of a truck.
This was a design study for a general purpose fridge for 2-inch cased holes (38 mm maximum OD) that would solve many of the temperature problems of downhole operation for a wide variety of sensors.
Use of my laser noise canceller to achieve shot-noise limited performance in SPR spectroscopy (in cooperation with Xi Wang and Andre Knoesen of UC Davis, and a bunch of my friends at IBM Almaden Research Center).
This was a photon budget for an OCT system—interesting primarily for the effect of path delay in turning FM noise in the superluminescent diode (SLD) into AM noise in the measurement.