Shorting negative and ground on PSU for cleaner signaling

[anythingthatworks]

I’m having communication issues that seems very similar to:

USB reliability when in AXIS_STATE_CLOSED_LOOP_CONTROL

Hi, thanks for an awesome piece of hardware! Unfortunately I am experiencing USB connectivity issues but only when the motor is driving, or even just holding a position with no load on it. For example: USB is stable until I enter command: odrv0.axis0.requested_state = AXIS_STATE_CLOSED_LOOP_CONTROL At which point the Jetson TX2 will repeatedly disconnect and reconnect (increasing the device number each time) until it gives up. Device fails to enumerate. Also tested on two windows machines …

And from a previous posts intermittent communication issues have been attributed to EMI. I went and got short USBs, ferrite chokes, USB isolators but without solving the issue and I’m hoping now it’s something I can solve by using CAN protocols? or with some grounding magic. Also according to:
Understanding the risk of a ground loop it seems like it might be ground loops but my Odrives are isolated so I don’t think it’s that.

This is all out of my league for now but the working theory is that the power supply noise is screwing with Odrive->computer communications because I’m on the latest firmware, and I get communication issues with the boards even if the motors are unplugged.

My question is will shorting the negative terminal and power supply ground help /be something worth checking or will that explode ?

[Riewert]

If you have communication problems without the motors on then there is something very strange/noisy.

Try running off of a battery, so you can determine whether it is the powersupply or the environment.

[PaulM]

I was running into these kind of problems and switched to CAN which has been working pretty darn good. A bit of a hassel to set up but once it is going it is a good standard.