So I had the pleasure of attending CES and seeing some of the VEX hardware. Discussion on the actual event can be seen here
What I want to talk about here is just the motors.
From @Paul Copioli we have heard to expect 2 - 2.5x the power output from a new V5 motor that we get out of a 393 motor. The internal DC motor is a lot beefier than the 393 internal motor so the 2x improvement makes sense. You can also see the 3 different gearing modules, the different colors are visible while inside the motors so no more confusion between different motor gears.
(photo credits to @RobotMesh )
The outer shell of a motor module is noticeably bigger than a 393 motor. A 393 motor with an IME is roughly the same width (?) as the new V5 motors but in the other 2 dimensions the V5 is larger. Hopefully needing less per mechanism makes up for it.
The plastic gear inside the motor was a question that came up in a previous thread. This gear is followed by 2 stages of planetary gear reduction so don't be too worried. While we will never know how the internal motors components will handle long term wear and tear, the only engineering solution is to make the whole system over built. Paul and I put the motor under insane load, back and forth as fast as we could by hand and the results blew me away.
The motor internal components survived several minutes of torture that would have stripped every gear in a 393. The axle we were using got very hot and twisted while the little gear showed no wear.
The "smart" aspect of the motor is quite interesting as well. Rough estimates put the chip inside each motor at being 50% as powerful as the chip running the entire current Cortex which is insane. The motors have their own diagnostic information that can be accessed on the V5 brain AND from Vex coding studio. This is game changing. Built in PID and motion profiling is supposed to all be customizable but as I was not programming anything I did not play around with this functionality. They even mentioned being able to change the update rate of the built in PID and that sending raw PWM values could be done a lot faster than 10 ms.