I have a stepper motor controller, PSU, and Nano Every mounted on an aluminum plate. PSU is case is grounded to the plate, as is the controller heat sink. Controller drives a stepper motor on my mill Z-axis. Motor cable is shielded. The Nano reads switches and a potentiometer for direction and speed. Cable to the switch assembly is shielded. Both motor and switch cables are about 3' long.
If I ground either shield, or neither, to the plate it behaves as expected.
But: If I ground both shields, lots of skipped steps. Basically non-functional.
Scratching my head.
Theories, ideas?
Thanks
That is no, no, no. Never connect chassis etc and ground.
The best would be if You post schematics especially the powering. Pen and paper works well, Fritzing pictures are not good.
1 Like
measure the voltage from GND to SHIELD.
1 Like
Shielded wire has only ONE end connected to your ground! Otherwise it is just another conductor that is inducing noise onto your signal wires.
1 Like
Thanks;
Not sure what you are suggesting. If only one end of the shield is connected, it is an open circuit.
The usual shielding rule is to not ground both ends, because the shield becomes a closed-circuit allowing circulating currents. The classic ground loop problem.
In my case, the shield at far end of both cables is not grounded. So the shields are open-circuited even if grounded at the chassis.
Something more complicated is going on. Maybe some antenna effects from a grounded shield?
No it is not. That becomes NO shield. That is just wrapping an unconnected wire around the others. Please provide links to such rules!
The shield should be grounded at one end , the end connected to the receiving device. Think of the shield as a long, skinny Faraday cage.
1 Like
Not to radiative (RF) noise, which it conducts to the appropriate ground, as it is designed to do.
1 Like
We (in my past career) had problems where there was differing ground potentials usually due to actual ground currents (as in rail or automotive), or poor mechanical connections between equipment. Grounding both ends just put all those ground currents into the shield, and then coupled into the circuit.
Grounding and Shielding Tips | Anaheim Automation
alphawire-Understanding-Shielded-Cable (1).pdf (312.7 KB)
On the other hand, the Siemens automation docs say ground the shield at both ends. Seems to be the rule for VSD and other higher current applications. Apparently, a single ended shield becomes an antenna.
Thanks again for your comments.
If you built your Arduino project inside a metal box or cage with unknown currents roaming around in the metal, like automotive or railroad locomotive, then it would be correct to minimize their effect by putting wires into what is in effect the same metal framework as the car or locomotive.
But you are not putting your Arduino project into such an environment.
Following up.
The 5v supply for Nano is a small dc/dc switcher operating from the 48 Vdc motor supply. After measuring and 'scoping around in my circuit, I found that this supply was not isolated -- as I assumed. So, Nano 'gnd' is shared with the motor supply negative 
Disconnect that supply, power Nano via separate USB supply and the mystery goes away. I can ground both shields without skipping steps, and all signals look a lot cleaner on the scope.
Now to replace that supply with a true isolated supply.
Thanks to all for the dialogs
1 Like
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.