Hi John, It sounds as if you have two AC lines coming into the machine. One for the controller and one for the VFD. The comments about a single star ground point are standard industry practice. One breaker at the panel which protects the wire to the machine. Another breaker inside the machine that protects the equipment.
The Earth (green) comes in and is connected to a welded stud with a ring terminal, not a spade and double nutted so it can't possibly come loose. >From that point your controller and VFD are run. Never ever rely on the frame to carry the AC Ground (Earth) from one section to another. DC grounds are isolated from Earth Ground. You can, connect the DC power supply minus to the Earth ground and it may even be a code requirement where you are. Be aware, some PCs do this inside their power supply as do other types of COTS hardware. And here is the problem you have to watch out for. Just because your DC volt meter reads continuity (Zero Ohms) between all the ground points doesn't mean that there is actually a path of lowest resistance following what you think it should. The moment you get into the switching power world which includes the VFD, Stepper Motors and Servos along with switching power supplies you get something called impedance. That's based on the DC resistance (usually low) and a combination of the wire inductance and capacitance and frequency. Depending on how things are wired and routed the impedance of the return signal of the VFD may be lower through the RS232 (RS485) shield than it is through the green wire or even the shield around the power cable. So just imagine the very noisy VFD signal returns through the communication shield because that has an impedance of 200 Ohms while the green wire has 1K. That's problem #1. There's two types of electrical noise. Electrical coupling and magnetic coupling. Electrical you shield against. Usually with the shield tied to a common earth somewhere. Magnetic coupling is the same thing that makes a transformer or a motor function. A rapidly changing magnetic field caused by rapidly changing current in a wire is coupled onto another wire that is in close proximity. Shielding doesn't protect against that. Twisted pairs do to a certain extent. The best protection is distance. That's why you don't run the VFD power or Servo power tightly tie wrapped to the encoder signal. Therefore put the AC power side for the VFD on one side of the cabinet and run the control signals as far away as possible. There are lots of books written on this subject and probably as many opinions on what to do. Ground shield at both ends. Ground at one end. Don't ground. Twist the wires. Don't twist the wires. Start with the Star Grounding. Make sure there aren't any DC paths that are unexpected as explained earlier. Then start looking at how the noisy signals might find their way back through alternate paths. Filters, Ferrites are all useful to block the signals. Always best to block at the source. John > -----Original Message----- > From: John Thornton [mailto:j...@gnipsel.com] > Sent: December-20-15 10:27 AM > To: Enhanced Machine Controller (EMC) > Subject: Re: [Emc-users] Grounding Issues > > > I have a line reactor for the VFD but I've never hooked it up. > > http://www.automationdirect.com/adc/Shopping/Catalog/Drives/AC_Drive > _%28VFD%29_Spare_Parts_-a-_Accessories/AC_Line_Reactors/LR-23P0-1PH > > Speaking of VFD's mine is grounded to the breaker panel and not in the > machine anywhere. When I check from the ground terminal to the panel I > get 0 ohms so I assume that the back of the VFD is grounded to the > ground lug on the VFD. > > JT ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
