Wireless problem - losing communication

[jalburty]

When using the joysticks to operate under a wireless connection we are consistently losing communication with the robot whenever suddenly going from full forward to reverse. It happens in arcade mode or in tank mode (only one joystick going from full forward to reverse in tank mode does not cause the problem). The robot then comes to a stop with the LEDs flashing on all four Victors indicating no signal and the driver station indicates no communication.

-We can regain communication by switching the driver station to disabled and then back to enabled and waiting about a minute.
-The problem happens just at the point when the motors start to reverse.
-The problem happens with the robot either up on blocks or not and we have checked our electrical contacts very carefully.
-The robot works fine in tethered mode.
-The problem does not happen if we go from forward to reverse slowly and gently.
-We are using the C++ version of the default code.
-We are using four Victor 884 motor controllers.

I suspect either the wireless router or the wireless bridge is the problem. Perhaps too much data is being generated by the joysticks for the wireless link to handle. The problem occurs only when going from full forward to reverse somewhat suddenly so we are able to get along for now, but this must not happen during competition.

Anybody else witnessed this behavior?

Joe Alburty
Team 1763

[BornaE]

Are you using a robot battery or a power supply?
if you are using a power supply, the sudden spike in current lowers the voltage to a point which the wifi adaptor or the crio resets.

if you are using a battery, make sure your battery is fine and the voltage does not drop too much.
either way its a power loss issue

[EricVanWyk]

Please verify that the bridge and cRIO are plugged into their dedicated connectors on the PD, and that nothing else is sharing their power.

If the bridge isn't powered by its dedicated power supply, it can drop out and create a situation like the one you described. Next time this happens, watch the lights on the bridge. If it goes through the same sequence as it does during power on, you have verified that the bridge isn't getting good enough power.

* Below this point is a ramble. Ignore or Enjoy at your leisure. *


Slamming the motors from one direction to the other draws a *lot* of current, and can be the hardest thing for the battery. The peak draw can approach twice the stall current of the motor in the worst case scenario.

Lets simplify to a perfect battery driving a motor that is a resistor and a back emf in series - ignore the series inductance for a moment. At full speed forward at steady state, the motor will have a back-emf that is equal to the battery voltage: No torque is being applied. When you suddenly slam the motor in reverse, this back-emf now adds with the applied voltage. For a brief moment, it is almost like driving the motor with a 24V battery. In this idealized case, the peak current is exactly 2x the 12V stall current.

When you go non-ideal, several things reduce this current. For one, the losses in your drive train will force the motor to apply torque to keep at steady state. This torque translates to a current, which translates to a voltage that develops across all of the resistances in the circuit. This reduces peak back-emf. Additionally, the inductance we previously ignored won't allow the current to instantaneously change, which will further reduce the peak.

In either case, this massive current slug is relatively quick - it only lasts as long as the motor is still spinning the wrong way, and the motors are being given a lot of power to convince them to spin the other way. This is therefore a quick but serious event that can easily offend the delicate sensibilities of the bridge.

The boost supplies on the PD are designed to handle this sort of thing and should be just fine.

[jalburty]

Thanks for the excellent and quick replies..

-We are using a battery and it does drop from above 12V to below 9V during the reversal operation.
-I don't think we are using a dedicated connector on the PD for the bridge - just a 30A breaker terminal.
-We are using the 24V dedicated connector for the cRIO.

I'll check these items out again on January 3rd when I can get back into the room. Perhaps I'll re-plug the bridge into its 12V wall wart next if needed.

Joe
Team 1763

[EricVanWyk]

Originally Posted by jalburty

-I don't think we are using a dedicated connector on the PD for the bridge - just a 30A breaker terminal.

There is the issue. Swap that over to the dedicated port and let us know if you still have issues.

[Hadi379]

We are also having this same problem. Our wireless bridge is connected to the dedicated 12v source. Not sure what to do.....

[Al Skierkiewicz]

Hadi,
I have to ask if you charged the batteries before you used them from the KOP? Then working down the list, check that all of the terminals are tight and that you don't have a damaged 120 amp breaker. You have used the #6 wire that came in the kit for wiring from the battery to the PD and the main breaker. Then check that all the screws are tight in the connector you are using to feed both the OI access point and the robot wireless. A quick check it to measure the output of one of the WAGO connectors with a breaker installed. The voltage should not vary much from 12 volts. If it does fall at the same time, as the router drops out, something is loose up stream. If everything checks out, and the CRIO voltage stays at 24 volts, then there is a possiblity of a defect on the PD.

[popo308]

We were having a similar issue...
Our problem was when we suddenly changed directions the power draw was causing the Gaming adapter to lose power momentarily...

The Fix? A fresh battery... now it works fine! Whenever it starts to happen we just swap the battery out for a new one and the problem stops...

[Hadi379]

All connections seemed to be secure and wiring is correct (gauges and connections). The battery was a fresh battery. I will check the output tomorrow. I am using an older 120amp circuit because one was not included in the early ship of the control system. I will also switch that tomorrow.

This problem seems to only happen when we make the sudden change from one extreme to the next, full forward to full reverse.

If this is a battery issue, what will happen in the middle of a match or near the end when the battery becomes drained a bit?

Thanks for the advice, hope it solves our problem...

[Al Skierkiewicz]

Hadi,
The power supplies built into the PD should work even with heavy battery draw. Remembering that any series resistance will drop voltage as the load current increases, teams should check for loose hardware on the battery terminals, PD terminals, loose crimps on the Anderson plugs and damaged contacts on connectors. It is quite common for teams that use the supplied alligator clips on the battery chargers to scratch the contact surface of the Anderson connectors and raise the contact series resistance. It is highly recommended (read that as do it) to change the clips to an Anderson connector. There is always the possibility that the battery is defective as well. Checking with a voltmeter will point you to the problem I think.