Crazy voltages on 120/208 systrm

[Atxrl1983]

Ok so this is what I have. Simple 480/277 panel feeds trans which feeds a 120/208 LV panel. 3phase. All breakers off I get 120v hot to ground and 120v hot to neutral on feeders. All breakers on I get 110v a-phase to ground, 124v b-phase to ground, and 135v c-phase to ground to feeders. I have receptacles reading 120v hot to neutral. And 208v neutral to ground. I found this today. Did not have time or ladders to really look into it. Nothing is tripped. But 2 power surges were burnt. I'm actually excited to see what causes this. It is in the office of the GC of a new 19 story I am running. Still a hole in the ground. I wasn't sure where to post this. Let me know what y'all think.

 

[Gene B]

I don't think any single wiring error could cause that. Number one, you don't have a valid equipment ground. If you did, a breaker would trip and you wouldn't be exposed to 208v. Don't connect any 3 prong devices until fixed. Actually, you could get zapped just touching the cover screw (or the cover if it's metal).

 

[GoldDigger]

It is also possible that you do not really have all three phases present. You can double check that by measuring L-L as well as L-N.

 

[mpoulton]

You have some kind of neutral-ground bonding problem, plus at least one other fault I believe. Re-check all the connections from the transformer secondary to the 120/208V panel bus, especially including the neutral and grounding conductor connections. Also double-check the identity of each transformer terminal to make sure they are actually labelled correctly.

 

[JFletcher]

Ok so this is what I have. Simple 480/277 panel feeds trans which feeds a 120/208 LV panel. 3phase. All breakers off I get 120v hot to ground and 120v hot to neutral on feeders. All breakers on I get 110v a-phase to ground, 124v b-phase to ground, and 135v c-phase to ground to feeders. I have receptacles reading 120v hot to neutral. And 208v neutral to ground. I found this today. Did not have time or ladders to really look into it. Nothing is tripped. But 2 power surges were burnt. I'm actually excited to see what causes this. It is in the office of the GC of a new 19 story I am running. Still a hole in the ground. I wasn't sure where to post this. Let me know what y'all think.

Sounds like you might have a loaded MWBC with an open neutral fed from AC phase.

 

[Cow]

You have some kind of neutral-ground bonding problem, plus at least one other fault I believe. Re-check all the connections from the transformer secondary to the 120/208V panel bus, especially including the neutral and grounding conductor connections. Also double-check the identity of each transformer terminal to make sure they are actually labelled correctly.

I'm with this guy. Check N-G bond and verify wires are landed on the proper terminals. Oh, and I'd shut the transformer down until you figure this out!!

 

[Atxrl1983]

More info

More info

Ok I read replies and agree. But it is still on. Ok so yes there is no ground wire in the conduit. That needs to be fixed. I put a pigtail to the four square at the receptacles and itdid not trip. Background. The GC has been in this office for a year. Servers are connected to this. Everything has been working except 1 receptacle on a quad. That is the one burning the surge protectors. But I still read 208v neutral to ground on the all outlets. The servers work with this going on. I thought it would trip or burn the servers but it has been working for over a year. So the only change was the GC connected his job site radios and he got a little shock. He stopped using that 1 receptacle. The radios charged on the other 3 receptacles from the quad. This is going to keep me up thinking about it until I find the problem. just FYI I have 10 years I the trade and have my masters. It intrigues me. I will really get into all the trouble shooting this week and determine the problem. I will post what I find. Or should I leave it so the GC get a little shock. Haha. No I can't do that. I will figure it out and fix it

 

[Atxrl1983]

Oh yea

Oh yea

So at the panel all breakers off I get the correct voltage to the feeders. I tried turning on just one phase at a time (1,7,13ect. ) I still read wrong voltages. So the feeders and transformer is good. It goes bad once I turn on the branch circuits

 

[ActionDave]

There doesn't need to be a green wire if the conduit qualifies as an equipment ground.

The servers probably have switch mode power supplies so they would not care if the voltage was high or low.

There is an open neutral somewhere. I can't tell from your description if it is in a branch circuit or the whole secondary of the transformer.

Check that. I just saw your last post. Sounds like the problem is in the transformer or panel.

 

[Atxrl1983]

There doesn't need to be a green wire if the conduit qualifies as an equipment ground.

The servers probably have switch mode power supplies so they would not care if the voltage was high or low.

There is an open neutral somewhere. I can't tell from your description if it is in a branch circuit or the whole secondary of the transformer.

Check that. I just saw your last post. Sounds like the problem is in the transformer or panel.

I I will check the transformer. The voltages are good when the breakers are all off. And it has been working for a while. But yea I thought of a floating neutral and I will verify. I will post what I find

 

[JFletcher]

My guess as to why the eqpt is still working is that you have an almost balanced load across phases, which would explain the 110/135V readings, the 208V neutral to ground (open neutral). If you take a household system MWBC (120/240V), connect a 60W lightbulb on A leg, and a 60W bulb on B leg, and lift the neutral at the panel, the lights would continue to work just fine as they are balanced loads; instead of 120V to each bulb in parallel, you have 240V in series with identical current drop across each. If you were to measure neutral- ground, you'd get 240V. I believe that is what's happening here just with 3 phases.

As to the burnt power strips, any load plugged in would cause a more severe voltage imbalance than what you measured, and let the smoke out.

 

[ActionDave]

I I will check the transformer. The voltages are good when the breakers are all off......

That's because there is no loads upset the balance.

 

[GoldDigger]

So at the panel all breakers off I get the correct voltage to the feeders. I tried turning on just one phase at a time (1,7,13ect. ) I still read wrong voltages. So the feeders and transformer is good. It goes bad once I turn on the branch circuits

If all of the loads are off and the transformer wye point is not grounded but an effective neutral goes from the transformer to isolated neutral bus in the panel, then it is reasonable to expect balanced incidental capacitance to force the neutral point close to ground.
Once loads are connected a much higher leakage or an actual connection, such a might be part of a surge protector, would then move one of the points of the wye close to ground.
This would force the other two points of the wye to about 208 volts to ground.
To see 208 volts from neutral to ground, I can see no mechanism other than a reversal of hot and neutral on that receptacle.
The only observable bad effects would be on loads which had poor insulation resistance to ground or surge protection varistors which went from hot and neutral to ground.
And the resulting leakage would also possibly account for the mild shock.
It is possible that the receptacles are perfectly well grounded by the yoke and the raceway serving the box.
In summary:
1. One wiring reversal.
2. No connection between X0 of the transformer secondary and ground.

 

[gar]

160329-0815 EDT

If I assume a triangle with 3 sides 208 long, then the ideal neutral point is at the center of a circle of 120.089 radius.

Next if I draw 3 circles with radii equal to the measured voltages of 110, 135, and 124, and centered at each of the triangle intersection points, then I have three different intersections of each pair of circles near the center of the triangle. These three intersections form a circle of radius 4.95 . Using the center of this circle as the apparent neutral point, then the required voltages to fit into the triqngle would be 108.2, 131.8, and 121.5. Close to the measured values, but off by more than the likely accuracy of the meter. Meaning the 208s were not all 208, or the voltages changed as the meter was moved around..

The resulting vector from the true neutral to the apparent neutral is 13.6 V at 117 deg relative to one of the 208 vectors. A voltage measurement from the wye neutral to earth should be of some measurable value.

Clearly the earth ground and the wye neutral are not at the same potential, or the 208s are not all 208,

Depending upon the meter load and whatever the leakage paths to earth are may even influence the voltage readings when moving from one phase line to another for the voltage measurements.

From the information we have it does not look like the wye neutral is solidly connected to earth.

.

 

[don_resqcapt19]

Are all of the phase to neutral voltages 120? If so, I would expect that the system bonding jumper was never installed.

 

[Atxrl1983]

Found it

Found it

So transformer was good. We had 3 multiwire branch circuits in a conduit. Neutrals were crossed. Same phase on 1 neutral. Easy fix

 

[don_resqcapt19]

So transformer was good. We had 3 multiwire branch circuits in a conduit. Neutrals were crossed. Same phase on 1 neutral. Easy fix

That is unlikely to have been the cause for the voltages you were seeing. All that really would have done is increase the load on the neutral and cause a small increase in the voltage drop.

 

[Smart $]

The voltages are good when the breakers are all off.

When you check voltage, telling us it is okay does not offer us any chance to see a possible problem you may be overlooking.

A to B ____ (208V nominal)
B to C ____ (208V nominal)
C to A ____ (208V nominal)
A to N ____ (120V nominal)
B to N ____ (120V nominal)
C to N ____ (120V nominal)
A to G ____ (120V nominal)
B to G ____ (120V nominal)
C to G ____ (120V nominal)


All of the above when off, and then again when on.

 

[kwired]

When you check voltage, telling us it is okay does not offer us any chance to see a possible problem you may be overlooking.

A to B ____ (208V nominal)
B to C ____ (208V nominal)
C to A ____ (208V nominal)
A to N ____ (120V nominal)
B to N ____ (120V nominal)
C to N ____ (120V nominal)
A to G ____ (120V nominal)
B to G ____ (120V nominal)
C to G ____ (120V nominal)


All of the above when off, and then again when on.

Loaded/unloaded?? Hopefully they are all zero when off

 

[Smart $]

Loaded/unloaded?? Hopefully they are all zero when off

Agree with zero when off... on the load side of breaker(s). I was referring to measurements on the line side... wherever he is getting the "crazy voltages".