SalomonTrujillo August 2005 Notes

RiSE? power problems

An error sometimes occurs when activating the supervisor on the RiSE? platform. The issue occurs when the carrier board is instructed to cycle power on the MCUs (I'm assuming these are embedded microcontrollers. Someone correct me if I'm wrong.) The CPU will then reboot (not hang). So, this isn't a segmentation fault, or a crash. The current theory is that this is a power related problem:

• Fix attempt 1: use a power supply for instead of the batteries for the CPU power (motors still use batteries.)
  • Interesting notes: At ~16 volts, the CPU draws nearly 700 mA during start-up, and 400 mA at idle. About 500 mA during the supervisor and an additional 300 mA with the LSANS attached.
  • This has not resolved the problem.

• Attempts to try:
  • The issue might be in the power block. Disconnect it and use power supplys to provide 5v and 3.3v (These supplys will need to provide a considerable amount of current)

• Talking with DonCampbell? about RiSE? problems
• Don suggested:
  • Scoping the 3.3 volt supply to see if the voltage drops during the MCU power cycle (He suggested single capture to trigger on a falling edge at 3.0v)
  • Disconnecting the RMCs from the carrier (this is a fairly intense operation, so I think we should to this last.)

RiSE? Programming

I've included almost all of what I know into RiSEHelloWorldTutorial? , but I believe that we will need to understand how the RiSE? platform implements its control laws. It took me a long time to find that code. It is in the file:

~/RoboDevel/RiSE/RobotCode/Hardware/RiSE_RBHW/HipWriter.cc

-- SalomonTrujillo - 10 Aug 2005