Here's where my EE knowledge is a little low: I know in certain circuits,

Here’s where my EE knowledge is a little low:

I know in certain circuits, that absoulite ground is unimportant (digitally) and that introducing a ground can actually be bad as it’s different from an absolute level from that in the local components.

I’m prepping the wiring for IGentUS and the cable has a ground…Cool, I’ll ground the frame and the PSU…should I somehow additionally wrangle a ground to the RAMPS and Raspberry Pi?

If your power supply is grounded, they’ll likely have their grounds coupled through the mains ground.

It’ll be grounded…if I run a ground wire to it. :smiley:

(edit: and I intend to run a ground wire to it.)

I wouldn’t ground the RAMPS or the Pi to the frame. Just me. I once built a self balancing unicycle thing. All worked great, then I tried to minimize everything and grounded, I want to say the IMU and arduino to the frame. I created massive noise in both circuits. The frame became a giant magnetic field. I know, a complete different animal, but I’ll never ground an arduino again.

@Brandon_Satterfield I don’t remember when, but I experienced something similar (or read a warning about it). Which is why I asked.

The voltages involved in the PI and RAMPS aren’t that great (still wouldn’t want to lick the terminals), the real voltages are in the PSU, which WOULD be grounded to the frame, and to the local ground…in case of something stupid like a massive short…rather than making the operator the closest path to ground.

On Eustathios I used shielded cabling on the steppers and brought that to the power supply ground to help with noise. Not sure if did anything, but its the cabling I had and figured it might help.

In my previous build I didn’t use shielded cables for the motors, but that caused some false temperature readouts. Shielding / grouding solved that. For safety it is a must.

In general best practice is a ‘single point ground’ . When dealing with analog signals, serial communications and digital I/O shields and 0 volt references should be tied to a single point. Most communication errors and analog sense problems are a result of ground loops [potential differences between ground points]. Shields are really noise antennas taken directly to ground !