Hello, my name is Ken and i write on behalf of rooky team 3336 from Swansboro high School in NC. Our regional is in Raleigh is this weekend and we have recently discover a problem that will likely prevent us from competing and I hope I can find some help. I have reviewed reams of documentation and forums and thus far can not find a solution. Sorry that this is quite lengthy but I felt it critical to list all the facts as we know them. Help!
Here are the details:
Problem:
Randomly all I/O dies (changes to an off state) for a fraction of a second and the come back. This included PMW signals to the Jaguars, solenoid output from the cRIO, and relay output. This happens at random intervals and can range from every 10 seconds to up to 20 minutes.
Additional information:
We are running Java.
During the brief I/O loss periods the “Link” LED on the cRIO appears to change to solid (stop flashing).
Forcing the operation mode to autonomous to force the indicator lamps (robot signal lamp and the LED on the side car) to light solid reveals that during these interruptions the indicators begin to flash.
The user watch dog has been disabled for trouble shooting with no impact.
The latest mandatory Driver software (DSupdate1.1for2010.zip) and mandatory NI updates (FRClabviewUpdate2.1for 2010.zip) have been downloaded an installed (we think this is when the problem began). See question on version validation.
We have eliminated wireless communication as root cause by running with a crossover cable from the classmate to the cRIO.
The driver station “communications” indicator stays green during these I/O drop outs.
The cRIO imaging tool has a date of 2009, although the NI updates have been applied.
Driver-station reports NI version V20 and DS version 02.10.??.?? or 10.02.??.?? (can’t remember exactly but it appears to be Feb, 2010).
We are using the classmate for development and the driver station. The problem exists using both log-ons.
Power for all devices stays up during these I/O drop outs.
Robot code functions fine except during these I/O drop outs.
Conclusions:
Due to the indicator (on the side car) beginning to flash I believe that the system watchdog is triggering the shutdown of I/O.
Due to the “Link” LED on the cRIO changing to a solid state I believe we are momentarily loosing communications between the driver station and the cRIO long enough for the system watchdog to issue the shutdown of I/O.
These brief losses in communications are not long enough for the drive station to detect, react to, and refresh the screen thus the communication status stays green.
Questions:
Are my conclusions on track? If not what other possible issues might there be?
If my conclusions are correct:
What possible causes might be causing the loss of communications causing the system watchdog to trip?
I understand the system watchdog timeout value is controlled by the imaging of the cRIO and can not be changed. How do we validate the revision of the cRIO image, and what should it be?
The “Java Updates” document from the “First Robotics Resource center” states that the latest Java version requires the V20 NI update (we have). How do I verify the current Java revision and what should it be? If Java is a down rev, and we have V20 could this be the issue?
What is the current “Imagining tool” revision?
What is the current revision of the Driver-station as reported on the driver-station screen?
What might be root cause of communications issues (especially when running wired)?
What else am I missing?
Any additional help beyond answering the direct questions would also be appreciated. We are stuck, have run out of time and are desperate! Should this problem occur during competition we will have one spastic robot! Pleeeaasseeee! Thank you!
