Mysterious Ghost Signal on PWM 4

[Robostang 548]

Has anyone else had something like this happen? Our team had a unique experience last night while using NI LabVIEW. We played around with a control scheme for our drivetrain. It has four motors. We ran the program from the front panel so we could edit some parameters while the program was running. One of these was a toggle switch that controlled a case statement in the program. In the false case, each motor was set to zero (basically a disable switch even though we already have one) and the true case would set each motor to the value of the y-axis of our joystick. When the program ran, the toggle switch was false, and the robot did nothing, however, when the joystick was neutral and the toggle switch was flipped to true, all the motors were neutral except the one on PWM number 4. We tried changing it so it wouldn't take the joystick input and would just assign neutral to the motor like it did in the false case yet it still would run at full. We were baffled that the same exact neutral assignment block was working in one case and never in another.

Next, we tried making a new LabVIEW frc project. We deleted all parts of the code that referenced any sorta pwm output or created a drive reference. We then opened a jaguar on pwm4 and in the main loop of the program, gave it a neutral assignment. We ran the program only to have the wheel running at full speed as soon as the program executed. We tried moving the jaguar to pwm10 and controlling it and had no issues.

So now we've ruled out that the jaguar is the problem. The digital sidecar might be broken but I'm still convinced it has something to do with LabVIEW seeing as how in our first test it did obey the assignment we sent to it in the false case. I'm fairly certain my pc has the newest LabVIEW frc updates and that the robot has the newest image on it but I can't be sure. Could a mismatch in software updates be the cause of this problem. Anyone else experienced anything like this?