Gray Jaguar CAN closed-loop speed control

[Ether]

Was any team successful in running closed-loop speed control this season with Gray Jaguars with the optical encoder signals going directly to the Jaguars ?

If so, I have a LabVIEW programmer here who would like to work on this off-season because he never got a chance to do it this year, and he has some questions.

My understanding is the following:

- he will have to update the firmware on the jags

- he will have to use some sort of 3rd-party RS232-to-CAN converter bridge (because the Gray Jags do not have this functionality)

- there is no built-in support for the Jaguar CAN interface in the 2010 FRC LabVIEW installation, so he will have to use 3rd-party VIs from WPI located at http://firstforge.wpi.edu/sf/projects/canjaguar

- the 3rd-party WPI VIs do not support all the control and monitoring capabilities that the Jaguar offers (or so I am told)

Is the above correct? Are there any other issues to be aware of?

Finally, is there a technical document for the Gray Jaguars that explains the Jag's API in sufficient detail that he could develop his own VI if he wanted?

Thank you.


~

[timmmoore]

We used 1 black jag for rs232/can and encoder, 3 greys with encoder and 1 grey using limit switches. The 1 black and 3 greys were for a mecanum drive, the 4th was for our arm.
1. Update firmware to 90 or 91 on both black/grey
2. we used 1 black jag as the interface
3. We used version 39 from WPI + mods
4. version 39 has a lot more monitoring, see the speed control tuning vi and dashboard. We used current, limit switches and speed monitoring back to our dashboard for testing/debug
5. If you want to use mecanum you will need to mod the CAN mecanum for speed control - not that difficult or I can find our changes.
For some reason in speed mode the motor invert on the open isn't passed through, we modified the vi to handle the invert booleans

[Ether]

Originally Posted by timmmoore

we used 1 black jag as the interface

Are you aware of any teams which did not have black jag and used a 3rd party interface successfully?

Originally Posted by timmmoore

If you want to use mecanum you will need to mod the CAN mecanum for speed control

Does the 3rd-party (WPI) CAN mecanum VI use the Jag's built-in PID speed control, or is that the mod you are talking about ?

Was your team's primary motivation for using CAN to get speed control?

If so, did you do this because open-loop voltage control did not give the driver satisfactory control of the robot in TeleOp?

Did you use the Jag's position control in autonomous for more accurate positioning?


~

[Radical Pi]

Originally Posted by Ether

Are you aware of any teams which did not have black jag and used a 3rd party interface successfully?

I know of many teams that used the 2CAN as an ethernet-CAN interface, nothing else was FRC-legal

[EricVanWyk]

Originally Posted by Ether

Are you aware of any teams which did not have black jag and used a 3rd party interface successfully?

~

I only got to go to one regional this year, but I saw a fair number of teams using CrossTheRoad's 2CAN.

[timmmoore]

Originally Posted by Ether

Are you aware of any teams which did not have black jag and used a 3rd party interface successfully?

~

We were the only team I saw at the Seattle regional using CAN.

Originally Posted by Ether

Does the 3rd-party (WPI) CAN mecanum VI use the Jag's built-in PID speed control, or is that the mod you are talking about ?
~

That was the mod I was talking about. We had code for both the original mecanum vi and the modified vi and we used the CRio USER1 switch to switch between them. This was because we have had relability issues with encoders in previous years and we wanted a way to try both.

Originally Posted by Ether

Was your team's primary motivation for using CAN to get speed control?
~

We wanted it for autonomous but we didn't get the time to test with and without speed control. We used both during the regional, a wire broke on one of the encoders and we switched speed control off. The feedback from the drivers was it drove better without speed control though that may have been because they were driving with the wiring problem before we found we had a problem.

Originally Posted by Ether

If so, did you do this because open-loop voltage control did not give the driver satisfactory control of the robot in TeleOp?
~

The drivers found open-loop was easier to drive, we havent had a chance to find out why.

Originally Posted by Ether

Did you use the Jag's position control in autonomous for more accurate positioning?
~

Never got to test that.

[timmmoore]

Here you go, this is the CAN version 39 from WPI with our extensions. The new files are:

CAN Jaguar Set Output Speed.vi
CAN Jaguar - Open4MotorSpeed.vi
CAN Jaguar - MecanumDrive_PolarSpeed.vi

we also updated CAN Jaguar HolonmicDrive.vi so PolarSpeed is another option within it.

Tim