Jaguars Blinking - Possible Control System Issue

[MikeDubreuil]

Hello,
I have previously posted this message to the Chief Delphi forum here. I was told to post onto the FIRST forum for additional support. It's also interesting to note that another person has experienced a similar issue here.

In a nutshell what I may have uncovered is a possible extremely serious issue which would prevent a team from competing in a match. It appears as a bug that causes the Digital Sidecar to not output a PWM signal.

Software:
I have applied the LabVIEW FRC Update. I made a "Basic Robot" project in LabVIEW and changed the Arcade Drive in TeleOp to Tank Drive. I've also set the autonomous conditional to allow it to run. LINK TO CODE

Problem Description:
I turned on the robot and the 2 jaguars were slowly blinking- I know "Loss of servo link." Except, the connections are tight and were previously working.

The DS status:
Team: 1337
Mode: Teleoperated
System: Enabled
Battery: 12.3 V
DS Rev: 2008-10-29a

cRIO:
Power LED Green

Digital Side Car:
RSL Solid Green (changes to blinking green when disable switch on DS is toggled)

Jaguars:
Slow orange blink in unison

I then systematically re-powered everything:

1. DS (issue still exhibited)
2. DS and Wireless Router (issue still exhibited)
3. Reboot robot (issue still exhibited)

After about 10 mins I went back to the robot.
A.) The RSL was slow blinking the Jaguars were slow blinking in unison.
B.) The DS said "System: Enabled" and "Battery: No Code"

I re-powered the robot and everything is working fine.

[Joe Hershberger]

Can you enumerate what hardware you have plugged in where? PWM channels, DIO Module Slot, etc.

[Joe Hershberger]

Also, do you have the same code deployed to run on startup that is causing the blinking Jaguars?

[MikeDubreuil]

Yes, this code is deployed on the cRIO. A picture of the connections. NOTE: The wire gauge to the power distribution board is not FIRST competition legal.

cRIO Slot 4 -> NI 9403 -> Digital Side Car
PWM1 -> Jaguar -> "Left Drive Motor"
PWM2 -> Jaguar -> "Right Drive Motor"

[jross]

What LabVIEW update and cRIO image are you using?

Can you hook up to the serial console and see what is printed when the problem happens.

On the cRIO filesystem, in the root directory, there is a file called WPI_error.txt (or something very similar). Can you FTP to the cRIO and post that file?

[MikeDubreuil]

What LabVIEW update and cRIO image are you using?

LabVIEW Update 2. The LabVIEW image. I know there's a newer update out- but it doesn't say it fixes the issue below.

Can you hook up to the serial console and see what is printed when the problem happens.

I didn't have a null at the time. I can connect now.

On the cRIO filesystem, in the root directory, there is a file called WPI_error.txt (or something very similar). Can you FTP to the cRIO and post that file?

I have about a 3MB error file. It appears as though the time is in GMT + 2 hours. Around the time I had the issue, I see the following error thousands of time:

ERROR Code: -44036 occurred at 01:00:21 01/26/2009
Description: Invalid set operation. PWM channel has not been allocated.
Call Chain: Basic Robot Main.vi >> RobotDrive.lvlib:TankDrive.vi >> RobotDrive.lvlib:Motors.vi >> MotorControl.lvlib:SetSpeed.vi >> PWM.lvlib:SetValue.vi

EDIT:
When you did into the VI you find this funny comment, "Assumes that anyone accessing the error
is going to report it."

[Joe Hershberger]

According to your errors, it would seem that you did not call open on your drive train to come up with the reference. Either that, or somewhere in your code the reference to your drive is being replaced with an empty reference. This can happen when the big robot data structure gets passed through a VI, but you neglect to wire it through. It then gets a default value which is invalid and will cause this error.

[MikeDubreuil]

According to your errors, it would seem that you did not call open on your drive train to come up with the reference. Either that, or somewhere in your code the reference to your drive is being replaced with an empty reference. This can happen when the big robot data structure gets passed through a VI, but you neglect to wire it through. It then gets a default value which is invalid and will cause this error.

Hi Joe,
That is concerning to me because my code is the basic robot project with (to me) minor changes. If what you say is true than the default basic robot project comes with an error. Is it possible for yourself and the experts at NI to track this down?

[Alan Anderson]

Hi Joe,
That is concerning to me because my code is the basic robot project with (to me) minor changes.

Was one of your changes to replace arcade drive with tank drive? It might be helpful for you to post either a screen shot of that part of the code, or the VI itself, so someone can look it over for problems.

The problem I'd be looking for first is what Joe mentioned: making sure the drivetrain reference from the Open VI is wired properly to the TankDrive VI.

[Mark McLeod]

I took a look at your code and while your Tank Drive conversion is incomplete, your drive motor reference seems okay.

These won't affect your problem:
Your tank drive should use two joysticks instead of 1, and what is Axis 4?
Replace the use of Arcade Drive in "TeleOp Reset" with Tank Drive.

By inspection only mind you, it seems that in
-- autonomous it should run 4 loops, reset to 0 then start blinking the RSL
-- in Teleop, joystick 1 y-axis should move PWM 1

I'll try to get your code on a cRIO and repeat the errors you're seeing.

P.S.
I assume from your description that you are using Deploy, not Run.
I just ask because you'd get "No Code" whenever your PC stopped execution, became disconnected from the cRIO, or lost communications.