On 01/02/2016 12:24 AM, Gene Heskett wrote: > this is a single phase system delivering 240+ volts AC > to the two ends of a single core transformers primary winding.
Well, yes and no. It depends on what you are looking at. The point in normal AC systems is that your ground potential is at or near neutral potential. In the system we have here, we have no neutral and ground is at half the potential of the input. > While I personally am not 100% happy with a no neutral hook up, it is > legal as long as it never leaves the confines of the load machine. The > machine itself is grounded well. The fact that the core/frame of the > control transformer isn't grounded probably enhances the capacitative > coupling, which might amount to something in the 100 pF neighborhood. > If the frame/core were also grounded, I would expect some of that > un-balance to go away and the total, added up, would be less than the > delivered voltage primarily because the measureing meter is going to > absorb some of it when hooked up to measure it, causing its own > imbalance. Some points here; I agree with the no-neutral is a problem assessment. However, that is something we have to live with. Capacitive coupling does not depend on being grounded or not. It is an inherent property of wires running close to each other, such as in a transformer. The capacitance is often a bit higher, but that does not matter too much for the argument. The unbalance as JT showed is measuring with the same multimeter at two points. The fact that two different values com out of that is enough to conclude that the imbalance is in the system-under-measurement and not caused by the multimeter. Otherwise you would see the same discrepancy for both measurements. I have not seen a multimeter that is so bad that it will throw a measurement off more that few percent. [snip] > The reason I wrote 240+ is that 127+127=254, which if Johns meter is a > decent one, he should be reading within say 2 volts of that from L1 to > L2. With no neutral, only the secondary of that control transformer is > to be considered the src of a 127 volt AC feed, and either end of it can > and should be tied to the well grounded machine frame, but not both of > course. Having it grounded gives a direct path to ground for the noises > the VFD might induce from close proximity. This is where we disagree. Whether it is 2x120V=240V or 2x127V=254V does not matter. The principle of re-referencing an un-referenced system in the way you suggest is wrong for the purpose because the installation does not leave the machine. If this was part of a house-installation, then I'd agree to re-referencing, but it is not. Also, the VFD's path to ground is quite a different one from the one after the transformer. A good grounding plan for the entire machine and a careful plan of /all/ other wiring is how you reduce the noise. There are many more gotchas than just one ground connection or a VFD. It is the connection plan as a whole that defines immunity and stability. -- Greetings Bertho (disclaimers are disclaimed) ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
