The problem is there are no low impedance ground available. The mains power cable ground is not low impedance at high frequency and the common mode voltage of inverters will couple a high frequency current into the ground cable. The trick usually used is to isolate logical ground and increase common mode resistance to reduce the ground current which also will lower the high frequency ground voltage.
>From the outside you machine have two or more power cables and one ground >cable. There will be a capacitance between each output phase from the >inverters and ground. This capacitance connected to ground is switched between >the lowest and highest input voltage which of course will make a high >frequency current flow into the high frequency non zero impedance ground. If you connect to a star ground inside your machine they will all bounce around with the same voltage. Nicklas Karlsson On Fri, 20 Mar 2015 17:54:23 +0100 Peter Blodow <[email protected]> wrote: > Przemek, the shield picks up stray energy from the suroundings. The > inside of it is a faraday cage and free of electric fields, which is > what we aim at. Now, as the shield has picked up the noise energy, > (re)converting electromagnetic field energy to real voltage and current, > it must get rid of it somewhere. This is most quickly done by a low > resistance, i.e. low impedance, shorting the noise current to ground on > the sending side without reaching the signal "consumer". > Sorry for answering late, I was on a short trip for the last days. > Peter > > > Am 18.03.2015 17:11, schrieb Przemek Klosowski: > > On Mon, Mar 16, 2015 at 11:34 AM, Peter Blodow <[email protected]> wrote: > >> Steve, > >> this is a good description of noise reduction by shielding. To make it > >> more exact, the shield should be grounded at the end where the lower > >> impedance is, mostly the signal source. > > This is opposite to what Steve said, which was to ground near the > > signals are consumed, which makes more sense to me because shield > > potentials, if any, have smaller chance of leaking through to the > > signal wires. > > >From the impedance point of view, I would also worry more about > > interference near high-impedance nodes rather than low-impedance > > nodes, just because smal currents result in larger voltages there. > > > > Can you summarize the rationale for your recommendation in the > > language of electromagnetics? > > > > > --- > Diese E-Mail wurde von Avast Antivirus-Software auf Viren geprüft. > http://www.avast.com > > > ------------------------------------------------------------------------------ > Dive into the World of Parallel Programming The Go Parallel Website, sponsored > by Intel and developed in partnership with Slashdot Media, is your hub for all > things parallel software development, from weekly thought leadership blogs to > news, videos, case studies, tutorials and more. Take a look and join the > conversation now. http://goparallel.sourceforge.net/ > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users ------------------------------------------------------------------------------ Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
