ECN Forum Forums Technical Discussion Electrical Theory and Applications Floating Ground - device being tested or scope?

Appreciate the response.

I found some information online that might help explain why I think the two readings on the secondary should add up to the voltage found between the two isolated conductors:

http://www.aptsources.com/resources/pdf/Floating%20Output.pdf



I might be way off with the above information but I believe that the diagram could represent the secondary of my isolation transformer where the L1 and Neutral (or maybe it should be called "L2" since it is no longer connected to earth ground) are isolated from ground. The ground connects to the shield of the transformer, the chassis bond, and ultimately to the earth ground.

In my test there was wasn't a load on the secondary. I simple took readings with my DMM at the female cord end that is connected to the secondary.

Here's an exact image of the transformer I am using:



As you can see, the male end plugs into a standard 120 V outlet which feeds the primary. The female cord end comes off the secondary.

EDIT: Whoops - that looks like a 240 V male end. Okay, that isn't the exact same transformer. Swap out the 240 V for 120 V and everything else is identical.

More images from the previously posted link:

Just when I couldn't get anymore confused...

This situation has been bugging me all day. Moments ago I went back out into the garage and plugged in my Hammond isolation transformer and took measurements off the secondary with the DMM again. Same results. 120 V from L1 to N, 72 V from L1 to ground, and 20 V from N to ground.

Next, I got out my old Weston Equipment analog voltmeter and took the same measurements again. This time I only got 120 V from L1 to N. The other measurements, L1 to ground and N to ground, failed to move the needle. 0 volts both times! To make sure the meter was responding correctly I unplugged the transformer and took measurements at the wall outlet. Everything checked-out as you'd expect - 120 V from L1 to N, 120 V from L1 to ground, and 0 V from N to ground.

In your friend's case, he's probably looking at a floating system with ground detection on it. The ground detector forces a high impedance ground reference at the center point. In your case, I wonder if they haven't built in a high megohm resistance between their Neutral and case ground. Or some TVS leakage or capacitive coupling between Neutral and Ground. You don't want the two to add up.
Joe

When you are using a digital meter on anything that is "floating" it is going to give you strange results.
You see the same thing if you hook up to a wire that is just floating in a wall.

We had ungrounded isolation transformer driven "convenience outlets" in most of the larger computer equipment. It was deemed to be safer because you would not be shocked if you got across either leg referenced to ground.

I was thinking something similar, that a Filter was responsible for the readings; however it could simply be Capacitive Coupling, as the OP has indicated the Analog Meter's Pointer indicated 0 Volts on the L-G tests, yet the DMM measured E on both L-G tests.

Even though both Meters exhibit a High Input Impedance, the Analog Meter might be 100K Ohm, and the DMM might be 50M Ohm - much higher Z than the Analog Meter.

--Scott (EE)

My friend is working in an OR theatre where there is an isolation transformer for each theatre and yes, there is ground current detection monitors in each theatre, as well. He claims his tests add up to 120 V when metering the separate conductors from each to ground. He might be incorrect.

In the article I posted above the images show each conductor metered separately which adds to 120 V. I realize that I might be confused and that the article has nothing to do with either his or my situation.

For my transformer you suggest that you don't want the voltages to add up to the supply voltage. Why?

Your friend is correct. If he disconnects his ground detector(s), he should get different results, ideally zero volts.
You don't want them to add up because there is not supposed to be any connection to ground. I don't really like the terms, "phantom" or "ghost" when discussing voltages. We did have a ghost at WCLQ-TV in Cleveland, but that's a completely different story. My original saying is, "All the world's a voltage divider." Your DMM probably has an input impedance of around 10 Megohm. A VOM could be around 50Kohm/volt. For kicks and giggles, connect your VOM and DMM, both on AC volts, in series across your transformer. The voltages that you observe on both, should reflect the ratio of their input impedances. You might not see much on your Weston, while observing almost full voltage on your DMM.
So if you hook your 10 Mohm DMM up to one leg and ground, and read about 60 volts, you are probably dealing with about a 10 Mohm combination of resistive leakage, and/or capacitive, and/or inductive coupling.
Joe

I think I should clarify a couple of things. By TVS, I mean Transient Voltage Suppression. Floating supplies can be a wonderful thing, but not if they float too far away from Earth reference. Various forms of coupling can steer them to potentially deadly differences WRT ground. For this reason, it's not unusual to use TVSs, to clamp at least the unreferenced floating Neutral to ground. Let's say you use a 1.5KE51CA, between Isolated Neutral and Ground. It would start conducting around 51 VAC, and clamp your supply. But below that point, there is a leakage current that must be considered.
I just went to the Digi-Key site and pulled the datasheet for the Littelfuse parts. The PDF was too large to attach but it shows that the 51 volt part might have a leakage current of up to a 1 uA. That's not enough to cause a problem, but plenty to confuse an electrician with a Fluke.
Joe