Shared Neutral across panels, and derived from different sources.

[McCElectric]

Chasing down issues in new addition to building related to grounding, and harmonic issues. Discovered the Emergency lighting panel (ELP) that is fed from a panel in original building thats connected to back up power, has some neutrals tied together with our normal lighting panel (L1). This was discovered by reading V,I,R of the MDS feeding the addition with the power off. Another test was then done between the panels, if you remove ELP phase A circuit 1 neutral, turn the circuit off and turn it back on, if it stayed on it was a culprit, then we did same process with LP circuits till ELP phase A circuit 1 turned of the lights. Then we investigated for the spot they could potentially be tied together. We located two spots and cleared 2 circuits from ELP that were tied to LP1. We had 6 drivers go bad by the time we found the shared neutrals also lost 3 smart switches. I originally thought maybe they acted like a MWBC and went into series and those drivers got smoked, but there was always one neutral landed during the test except when we would turn a circuit back on with no neutral. The engineers were convincing it would be like opening a switch, the lights just won’t work. My other theories are these drivers can be damaged with over voltage or even under voltage, or the drivers could short out to ground, due to the test.



ELP 3 phase 277/480



LP 3 phase 277/480



One spot was same phase

One spot was different phases

 

[synchro]

Rather than disconnecting neutrals a suggestion is to (within a panel) put a clamp meter around both the line conductor and the neutral conductor of a circuit, and then verify that there's a negligible current reading when that circuit is turned either ON or OFF. If there is no connection from the neutral conductor of that circuit to the neutral conductor of another circuit, then all of the line current from the breaker on that circuit should return on its own neutral conductor. And so there should be a ~ zero current reading on the clamp meter. If you get a significant current reading then you can go from there to track down where neutrals might be connected.

Is it possible that there could be a connection between the neutral conductor of a circuit in one panel and a neutral conductor of a circuit in the other panel that's on a different phase, but the circuits are only connected by one conductor terminated to a neutral bus? If so, and you lifted that lone connection to a neutral bus, then you could get an overvoltage on the circuit with the lightest load.

 

[ActionDave]

Sounds like you lifted a neutral on a circuit that didn't have a bootleg neutral from the other panel.

 

[McCElectric]

Rather than disconnecting neutrals a suggestion is to (within a panel) put a clamp meter around both the line conductor and the neutral conductor of a circuit, and then verify that there's a negligible current reading when that circuit is turned either ON or OFF. If there is no connection from the neutral conductor of that circuit to the neutral conductor of another circuit, then all of the line current from the breaker on that circuit should return on its own neutral conductor. And so there should be a ~ zero current reading on the clamp meter. If you get a significant current reading then you can go from there to track down where neutrals might be connected.

Is it possible that there could be a connection between the neutral conductor of a circuit in one panel and a neutral conductor of a circuit in the other panel that's on a different phase, but the circuits are only connected by one conductor terminated to a neutral bus? If so, and you lifted that lone connection to a neutral bus, then you could get an overvoltage on the circuit with the lightest load.

I will look into the method you suggested, all circuits have their own neutral at their respected panels.

 

[McCElectric]

Sounds like you lifted a neutral on a circuit that didn't have a bootleg neutral from the other panel.

I warned them of the dangers of floating a neutral, they said it didn't apply here it would be like turning a light switch off. I am still questioning what the effect of the damaged drivers was from over voltage or some kind of short. The circuit that suffered the damaged lights, was the circuit with the bootleg neutral. after locating the circuits we (floated the neutral at another point) to see if the lights ahead or behind us would stay on during that test thats when I saw flickering and a possible pop. We found the location and took it apart, ELP worked fine LP lights up until the spot we disconnected were fried. We thought we had another issue until testing inside the ballast showed a drop of 277v.

 

[hillbilly1]

A live circuit tracer is the best way to find neutrals parraleled or fed from the wrong panel. Clip onto the neutral and hot from the farthest circuit. The neutral current will show up at the originating panel, or if paralleled, the signal will be split between the two panels.

 

[Fred B]

Sounds like you had a shared neutral on 2 or more circuits that once lifted created a situation typical of any dropped neutral where on phase or leg will bump high and other will drop low. That will burn out electronics quickly.
Neutrals of the associated ungrounded conductor of the circuit must always land in the same panel 210.4(A)(B).
See this trouble most time when someone that is not competent to work on the systems under some simplified (not always applicable) impression that tie all neutrals together in the jbox, but when they actually happen to be to 2 or more seperate circuits, a 200.4(B) requirement issue and is not followed creates a real issue.
When the circuit neutral becomes that intermixed and multiple connection best to shut down panel(s) and trace rather than starting to disconnecting neutrals.

 

[McCElectric]

Sounds like you had a shared neutral on 2 or more circuits that once lifted created a situation typical of any dropped neutral where on phase or leg will bump high and other will drop low. That will burn out electronics quickly.
Neutrals of the associated ungrounded conductor of the circuit must always land in the same panel 210.4(A)(B).
See this trouble most time when someone that is not competent to work on the systems under some simplified (not always applicable) impression that tie all neutrals together in the jbox, but when they actually happen to be to 2 or more seperate circuits, a 200.4(B) requirement issue and is not followed creates a real issue.
When the circuit neutral becomes that intermixed and multiple connection best to shut down panel(s) and trace rather than starting to disconnecting neutrals.

Yes one of the j boxes got tied together, we never opened the neutral with a load on it, all circuits had their own neutral. A couple were tied together, but still eventually went to their own panel.