Quote:
Originally Posted by phargo#1018
I am primarily interested in the local loop closure aspects of the new Talon. When using the Talon to close a motor control loop locally, what are the loop(s) that are closed?
Is the loop closure the motor rate loop, or is there a motor current loop as well?
Can you provide a list of parameters that are settable within the Talon?
Are you able to clarify the auto-disable/auto-enable features that may be associated with the new Voltage brown-out protection?
Is there a master/slave feature such that a pair of Talons can operate a pair of motors on a common gearbox?
Are you able to comment upon the device time response? I have heard that the Talon provides a 1msec output update. This sounds faster than the basic PWM command input, so I am trying to track this down. Is the update rate faster/slower when using the CAN interface?
Appreciate any answers!! Thanks!!!
|
Hmm... OK, let me see what I can answer... The Talon SRX is meant to provide all of the services from the CANJaguar class in WPILib. Therefore, you can close the loop on speed, current, position, voltage, etc. given the correct sensors. The parameters are similar to the CANJaguar class as well.
The brown-out protection will take you down to about 5-6V before the Talon resets. The motors will be dead likely before that happens.
In PWM mode, the talons can be tied to the same PWM output to slave them together. However, I don't see anything in the API that allows you to do that in CAN mode. We've simply been sending the same command to the different Talons via CAN. We haven't been able to detect much of a lag if any.
The update rate on the Talon SRX is indeed roughly 1ms. The bus is running at 1 Mbps which does make it faster than the PWM. However, your driver station is still running at 50ms refresh rate. So, in a totally closed loop mode, expect ~ 1ms response times. If you're taking commands from a DS, then the update rate is slaved to the refresh on the DS.
The docs on the Talon SRX API are still sketchy. Lots of things defined that the details are still a work-in-progress as far as the docs go. And, even the beta teams won't be able to see the implementation until after kick-off. Of course, I can only speak to the C++/Java WPILib implementation. What happens in LabView should be similar, but I can't say definitively.
HTH