I have never seen exactly what you have described in a 480V application. If the relays are the same part number, the coil voltage should be roughly equally divided and neither will likely pick.
Here is a setup that you will often encounter: A 480V to 120V control transformer will be powered from, let's say, phase A & phase B. Two 480V coil relays will be connected between A & C, and B & C, in order to check for phase loss. The NOHC (normally open held closed)"a" contacts of the two relays are placed in series to disable the control ladder should you lose a phase.
This method does NOT protect you from phase reversals or offer tight voltage tolerances. Phase loss monitors are more commonly used for this purpose in the present day. Joe
Thanks Joe. What I am after is field engineering a ground detector for a 480V ungrounded delta 3Ph equipment feeder. I just need to trip a alarm circuit in case of a phase ground with out having to spend a gazzillion dollars for a manufactured unit.
It sounds quite feasible. Series-connected relay coils arranged so that there wasn't enough current to operate either until some other condition was met were quite common in old telephone switching equipment.
If you connected a relay coil from each phase to ground, then a ground fault on one phase would result in the two relays for the "good" phases being operated. N/O or N/C contacts could then be wired to provide whatever form of signaling is needed.
I see what you're looking for now Bman. It's just that you wouldn't be using relays in series to do what you need. One bummer is that it is hard to sense for a big ground fault, without causing a little ground fault of your own in your detection equipment.
So you can sense the corners to ground as Paul has described. But think about another possible method... What if you connect a properly sized and rated fuse and series resistance from each corner into a common node in the middle. Then use any of a number of methods to sense between this node and ground. Choices here are driven by how high a value of ground resistance you are trying to detect as well as how much ground leakage you are willing to impose with your sensing circuit. Joe
Ok, thanks guys for the input. I'll try not to be redundant. My objective on this circuit is to simply detect an all out ground fault such as a phase wire getting pinched and grounded by a cover screw or such. Just don’t want to have the system operating as such and then have say another phase go to ground due to a device failure and powwweee. Not sure where or what to look for as to your scenario Joe other than to say I am familiar working with rail mounted cube relays and terminal blocks. Does the arrangement Pauluk describe (what I was originally thinking) sound applicable for my criteria? Thnx bman
There bman, you narrowed it down. The relays, to ground not in series, will do what you need to look for a hard fault. If you were looking for varying amounts of leakage, maybe not. Of course, proper fusing and protection, and a means of function testing, are very important. Mayhaps, the manufactured, tested, and listed, unit would cost less in the long run. Joe
Bman, it sounds like you are describing a ground-detector circuit applied to a 480V ungrounded system. If you want to build your own, a fairly common arrangement is to use three 100VA 480:120V control-power transformers with the 480V primaries connected in grounded-Y, and the secondaries connected in “broken delta.” The secondary resistor should be about 150 ohms/200watts [with plenty of ventilation], with a 208V-coil alarm relay connected to V4 and V5.