I'm an electrical contractor in Central California and I'd like to relay a strange situation I came across on a trouble call.
A large Hi Tech company called me in to one of their data centers to check out the reason for burning a 120-volt 2KVA UPS system after it was plugged into some recently installed receptacles. After verifying that all the effected loads were disconnected, I energized the receptacles on the new circuits, there were only (5) 120v branch circuits on a 225 amp, 3-phase 208/120Y panel, but several 208v single-phase loads. None of the circuits were multi-wire branch type.
I read the Line to Neutral voltage and it was 115.6, 117.8 and 117.6 on A, B and C phases respectively. Line to Ground was 118.1 on all phases +/- 0.1, and 207 volts between all phases. I suspected a Neutral problem but had no idea what I was about to find. I checked the voltage at the panel mains and found the exact same readings. All the neutrals were tight, including the main. So what happened? I traced the main neutral back to source, a PDU in the room fed by an 80KVA UPS. Within the PDU I found a 100-amp feed to the panel in question, a good ground but the neutral was never hooked up. It was safed off during install, tucked away and forgotten. Without the neutral in the panel, 208v was able to find it's way through the loads when they were plugged in and burn power supplies.
The real question is, where did my Fluke 87 get it's zero reference when I was checking the line to neutral voltage. I might mention that I megged the open neutral and it was >2,000 megohm at 250 volts, no grounds. To prove it I stuck a good old wiggy on each phase to neutral and it showed zero.
I got together with the on-site electrician and we tested several digital meters on the circuit and found each one had similar readings. I connected the main neutral in the PDU and all readings returned to normal and loads were connected successfully.
Is the moral of the story not to trust your Digital Meter? Where do you think it got it's zero reference? I don't have the answers and neither did an electrical engineer I spoke with. Any high impedance digital meter we put on the circuit showed voltages from Line to Neutral within spec, but low impedance equipment showed low or no voltage, phase to neutral.
This is a new one for me.

Bud


Bud