Attending: Tim, Andrew, Glenn, Jacob, Sarah, and various Software Team members.
We discussed the downsides of the STM USB ARM chip. We found quite a few "warts". We had issues with the clock and PLL systems, the internal voltage regulators, USB implementation, and ADC and oscillator slowness. The full dissection can be found here.
Since the STM chip has these warts, we started to explore other alternatives. The Philips CAN and USB parts were well documented and had good tool support. However, the CAN chip has 16KB RAM, which is below our 32KB requirement. Also, the USB chip uses an internal 1.8V regulator and the CAN chip does not. This means we can't just drop in a different chip like we can with the STM chips. Atmel has great USB support, but no CAN chips. Both vendor alternatives don't offer a drop-in solution, but we weren't happy with the STM chip.
It was decided to take another week to dissect the Philips and Atmel USB chips to see if they were any better than the STM chip.