Part Number: LAUNCHXL-F280049C I am running a smallish 3-phase motor ( Maxon EC-i 40 488607 ) using the LAUNCHXL-F280049C and BOOSTXL-3PhGaNInv . The BOOSTXL is powered by a linear 48V supply. The LAUNCHXL is powered from USB, connected to a PC (J5 on the BOOSTXL is un-jumpered so that it doesn't attempt to power the LAUNCHXL). I also use serial over the USB to enable/disable PWM output, control speed, and monitor phase currents using an application on the PC. The setup works if I use a laptop (not plugged into mains, just battery) instead of a desktop PC. I can start and stop the motor,adjust its speed, and monitor phase currents on the PC. The setup also works as expected using a desktop PC when the 48V supply is turned off. That is, the LAUNCHXL enables and varies PWM output as intended (though obviously the motor doesn't run because the BOOSTXL is unpowered). The setup fails using a desktop PC with the 48V supply turned on. Setting motor speed to a non-zero value starts the motor spinning but then the USB connection on the Windows PC fails with "USB device not recognized: The last USB device you connected to this computer malfunctioned, and Windows does not recognize it." Other interaction, such as enabling or disabling PWM or viewing phase currents, works until the motor speed is made non-zero. Once the error happens only disconnecting and reconnecting the USB cable (which removes LAUNCHXL power and so brakes the motor) will allow USB communication to be re-established. The power draw from the motor starting should be low, say 2A RMS, because I ramp up the duty cycle gently and there is minimal mechanical load. In any case, I suspect that the initial current-draw on the 48V supply when the motor starts is doing something peculiar to USB power rails, perhaps via the power ground on the BOOSTXL, which is connected to ground on the LAUNCHXL and then on to USB ground. USB ground is mains earth on the PC and I am guessing that the PC doesn't like it if some different "ground" voltage is connected, creating a significant amount of current. In contrast, on the battery-powered laptop the USB ground floats. (An alternative speculation is that the laptop has some more sophisticated response to oddness on the USB connection, such as automatically re-connecting after an internal trip, whereas the PC just dumps the connection for good and gives an error message). I have tried adding a functional earth, shown red in the image above, to pull the negative end of the 48V DC supply more strongly to mains earth, but that does not resolve the issue (there is about a metre of 2.0mm copper wire, insulated, from the 48V supply output to the mains earth pin). I have also contemplated splitting PGND and GND on the BOOSTXL board, but some of the circuitry is connected to both 48V and GND (e.g. phase current detection), so that a split ground could cause over- or under-voltage troubles if the PGND floats away. The USB isolation jumpers JP1, JP2 and JP3 on the LAUNCHXL are all jumpered, i.e. no USB power isolation. We don't currently have the ability to provide separate 3.3V or 5V supplies if the jumpers are removed, and I am concerned that removing JP1 on its own (isolate ground) risks problems like those already mentioned for splitting PGND and GND. Does anybody have any suggestions?
↧