Classmate Static Discharge Susceptibility

[charrisTTI]

We had two matches at the DC regional where we were unable to drive the robot even though all the communications led indicators were green.

In the first instance the joystick operating the kicking mechanism worked fine, but the joystick controlling the drive motors did nothing. After the match we returned to the pits and connected classmate to robot with a tether. Robot was restarted and driver station software was restarted (exit and relaunch). The system continued to exhibit the above behavior. Complete power down of classmate and restart restored functionality.

In the second instance both joysticks were non functional. In the pits after the match we attempted to exit and relaunch the driver station software, but the dashboard window would not close (something was hung up). We forced the classmate off by holding down the power button and restarted which restored full functionality.

I believe that static discharge to the Classmate/joysticks occured to cause these problems.

To mitigate these problems we:
Added a ground strap to driver station to give good connection to metal frame of field
Restarted classmate from power off state when we entered the queue
Instructed drive team to be sure to touch metal field frame before touching joysticks at start of teleop.

Drive team walking on the carpeted surface while waiting for the match to start and autonomous to finish can generate plenty of static electricity.

Other ideas??? Does this explain any sympotoms which you observerd at your event?

[Mark McLeod]

Similar symptoms I saw this weekend were tracable to under-powered USB devices. All Comms are good, but the robot doesn't respond to the driver controls.

These are the USB mistakes I saw teams making:

If you are using the PSOC it is usually the biggest power-hog. That must be plugged directly into the Classmate.
The USB hub should be plugged into the other Classmate USB connector and the joysticks into that.
If you have a real power-hungry game controller - lights/vibrate/etc. - that needs to be plugged directly into the Classmate too. So you may have to choose between the PSOC and any fancy game controller.
The Stop switch they gave us in the KOP is NOT used at competition. Unplug the power drain to run those color changing LEDs.It seems the hub is a major failure point and it's difficult to predict if a team will have trouble with it or not. Bring a different model to competition as a safety backup and be mentally prepared to repatch your USB connections. Face it, you'll have nothing better to do if your robot is non-functional...

[charrisTTI]

Similar symptoms I saw this weekend were tracable to under-powered USB devices. All Comms are good, but the robot doesn't respond to the driver controls.

These are the USB mistakes I saw teams making:

If you are using the PSOC it is usually the biggest power-hog. That must be plugged directly into the Classmate.
The USB hub should be plugged into the other Classmate USB connector and the joysticks into that.
If you have a real power-hungry game controller - lights/vibrate/etc. - that needs to be plugged directly into the Classmate too. So you may have to choose between the PSOC and any fancy game controller.
The Stop switch they gave us in the KOP is NOT used at competition. Unplug the power drain to run those color changing LEDs.

The National Instruments rep at DC said to be sure to plug the RED power only USB cable into one of the classmate ports and the BLACK USB cable into the remaining port. We then had one KOP joystick, one Logitech Extreme 3D Pro, and the Cypress module plugged into the hub. Stop button stayed in the pit area. Still saw problems with this configuration.

[Bharat Nain]

Similar symptoms I saw this weekend were tracable to under-powered USB devices. All Comms are good, but the robot doesn't respond to the driver controls.

These are the USB mistakes I saw teams making:

If you are using the PSOC it is usually the biggest power-hog. That must be plugged directly into the Classmate.
The USB hub should be plugged into the other Classmate USB connector and the joysticks into that.
If you have a real power-hungry game controller - lights/vibrate/etc. - that needs to be plugged directly into the Classmate too. So you may have to choose between the PSOC and any fancy game controller.
The Stop switch they gave us in the KOP is NOT used at competition. Unplug the power drain to run those color changing LEDs.

That is interesting because I saw a couple teams with that configuration at NJ who had problems. Instead, I suggested that they plug bot the USB hub connectors into the classmate and power the PSOC via the hub. That seemed to work.

On a side note, I am so glad we did not have to use the PSOC this year.

[Greg McKaskle]

Can teams who had issues with joysticks on the field give a brief description of how things were attached to the computer? What was plugged into the left side, what was plugged into the right?

Also, what was the setup procedure you followed? Did you plug in enet, then open the classmate, open the classmate, plug in enet, then plug in USB, etc?

As for static discharge issues, are there any other symptoms to point in this direction? Any reboot after a spark, any joysticks that worked until a spark?

Greg McKaskle

[Mark McLeod]

If you mean on the field...
The Classmates were already powered up and USBs installed when they arrived and connected to the field Ethernet cable.
Most Classmates were running all day. The Driver account must be logged out after a match and logged back in.

Lots of variations on what was plugged in and where.
I'll use one team as an example, Dustin or Nick can correct me since I wasn't there every time, but 816:

First had the Stop button directly connected to the left Classmate USB port. With the hub to the right port with three joysticks.
Second had the red hub USB connected to the left Classmate port and the black one to the right.
Third, directly connecting just the two drive joysticks corrected the problem, but left them without kicker control.The common factor seems to have been the HSB hub, and seems to be due to a faulty USB hub or hub cable.
I'll be carrying a spare USB hub cable or two at the next event I work to perform a quick to test when this happens again.

But the solution for teams in the middle of a match is probably to design your controls to have a backup alternative USB configuration that just uses two joysticks directly connected to the Classmate. And be mentally prepared to repatch on the fly.

[Greg McKaskle]

The reason I'm asking is to try and cluster the symptoms.

By the way, when this is in failure mode, does the diagnostics page show the joystick, and do the diagnostic LEDs show blue when joystick buttons are pressed? Do all of the USB devices go out, or was it sometimes the ones on the left, sometimes the ones on the right? Is it ever the case that two joysticks were plugged into the same side of the computer, and one stopped working, but the other continued working?

Greg McKaskle

[charrisTTI]

The reason I'm asking is to try and cluster the symptoms.

By the way, when this is in failure mode, does the diagnostics page show the joystick, and do the diagnostic LEDs show blue when joystick buttons are pressed? Do all of the USB devices go out, or was it sometimes the ones on the left, sometimes the ones on the right? Is it ever the case that two joysticks were plugged into the same side of the computer, and one stopped working, but the other continued working?

Greg McKaskle

In our first failure of this type we had USB hub plugged into right hand connector and nothing plugged into left connector. We had KOP joystick, Logitech Extreme 3D, and Cypress module plugged into the hub. KOP joystick continued to work. Logitech was ignored. I did not check diagnostic leds at the time. Both joysticks did show in the list of devices.

In our second failure of this type, same configuration as above but RED power connector from hub plugged into left connector, both joysticks did not work. Driver station software would not cleanly exit. Dashboard continued to be displayed until we forced power off.