5v brownout (different that RR supply brownout)

Technical Control System
1

We are getting a number of 5v brownout messages in our log file. I am told the correct interpretation is that it is the internal 5v rail browning out, not the robotrio (which continued with no obvious side effects). The roborio specs documenting current limits is here. https://decibel.ni.com/content/docs/DOC-30419.

The Roborio uses 5v for a number of items including the 5v PWM outputs, DI/O, Analog Input. We have a lot of motor controllers (10), 4 encoders, 3 light sensors and 1 ultrasonic sensor. We have carefully checked the 0.1" connectors for loose strands etc.

There appears to be no relationship between vibration and the messages (we dropped our robot from 6" a number of times with no messages).

Anyone with expert specific knowledge (maybe from NI) confirm

  1. Are we interpreting the message correctly
  2. Could we possibly be at the current limit of the 5v rail with the above configuration operating correctly or not likely ?
  3. Any obvious diagnostics ? I suppose we could measure the current on each I/O connector and sum them ?
2

If you are overloading the 5v rail, try using the WCP Spartan Sensor Board for PWM. It has its own power connection and the PWM ports on the board use that power, and not the RoboRio’s power.

You might also be able to use the power from the VRM for the DIO and AIO loads (instead of using +5v from the DIO/AIO connector, get +5v from the VRM).

3

The PWM and Relay ports are straight pass through from the roboRIO on all pins, the other sections are powered off the Spartan board supply.

4

Thanks for the clarification. I had assumed that with R54 allowing the PWM ports on only the Spartan board for Servos, that somehow the Spartan Board was powering the additional PWM ports. Otherwise, why wouldn’t the additional PWM ports be allowed for other MXP expansion boards?

5

The PWM ports on other boards have DIO 5V power on the power pins because they are from the MXP. The ones on the Spartan board aren’t from the MXP, they are just a pass through of the roboRIO’s 10 dedicated PWM pins so they have 6v servo power.

OP, have you opened the roboRIO to examine it for debris? That would be my primary suspect if the current draw of your devices isn’t changing. It sounds like you can’t reproduce a scenario where this triggers?

6

If you aren’t seeing any other rails brownout, then that means the 5V is almost certainly shorting somewhere on the robot or within the roboRIO.

Greg McKaskle

7

we did blow out the roborio and tape up all openings at the regional. At the end we removed it and opened it up and saw no issues like debris. We did find a slightly out of position pin on the Ethernet socket.

No amount of deliberate vibration could trigger the message. The robot was dropped 3 times from 6" high powered on, then driven on the field at lunch deliberately crashing into field walls. So its not a vibration triggered (or permanent) short. The robot and its sensors appears so far to operate normally in spite of the message.

I am hoping NI will come up with a list of every internal device internal & external that uses the 5v rail. Perhaps the Ethernet link uses it too. It would be nice to know if the pwm signal is generated off the 5v rail as we have 10 speed controls, any one could be putting an excess load on the signal line and hence the 5v rail due to some type of internal problem.