Several recent posts have asked questions about the responsiveness of both the NXT and 12V motors. You may find the attached post in a different forum useful http://www.robotc.net/forums/viewtopic.php?f=1&t=2045.

NOTE: I recall using a 7.2V VEX battery to control the HiTechnic controller and 12V motors. I did this for the measurements and wanted to use approximately same battery voltage for all three motor types (NXT, VEX and TETRIX 12V).

Without knowing the moment of inertia of the load, it seems premature to be shocked, but offhand, those numbers seem horrible, even for a toy.

Was that just the Tetrix motor and gearbox with nothing attached?

Have you looked to see how long from the program commanding a stop to the message being transmitted on I2C or a change on the output of the Motor Controller? (that would factor out the mechanical issues).

All measurements were done with unloaded motors.

I've mentioned previously the HiTechnic TETRIX motor controllers have a "speed ramping" function embedded in their (non-modifiable) firmware. I believe it adjusts speed a maximum of 10 "steps" per clock "tick". So:

• Starting with speed 100 and the motor is commanded to go to speed -100.
• On the next clock tick the speed is set to 90.
• On the next clock tick the speed is adjusted to 80.
• And so on until the motor speed gets to -100.

So it will take 10 clock "ticks" simply to get the speed to zero. And another 10 ticks to reach the final speed of -100.

Clock "ticks" are the internal processing in the HiTechnic controller. The firmware in the controller recognizes a speed change on a periodic basis which is either 25 or 50 milliseconds. I can't remember which is used.

I believe this is the root cause of the slow response from the 12V TETRIX motors.

Last year we did some experiments and made a motor function that checks whether the user has done a reverse direction.
If so, the function would stop the motor, wait 20 msec then start it in the opposite direction.
That seemed to give us a much more responsive system.