I think I can beat that:

I had a weird, general purpose auditorium -- whose structure was such -- and my bosses were such -- that the most important branch conductors were run the L O N G way around a not so small building. (35,000 sf)

At issue: two 277VAC critical lighting circuits that feed ganged switches at the FAR side of the basketball court/ auditorium. These were re-engineered by me to be on the exact same phase so that there was no NEC violation in the switch box. Standard 277 switches controlled the overhead HO T-5 lights. (At the time, the latest 'thing') They really kicked out the lumens.

Because of the loading -- and distance -- I upgraded the neutral returns to #8 on both 20 breakers.

Now, get this, half-way around the building it was necessary to splice both hots, both neutrals -- on their home run back to the panel. My dufus apprentice made up this simple box.

Well, not so simple: AFTER getting the 'hots' properly wired through, under my specific direction from the floor, I turned my head, and he made up both neutrals --

Naturally he wired BOTH line-side returns to each other -- and BOTH load-side returns to each other. Obviously, the massive light array had not return to the Service panel as a result of this.

So, instead of failing to fire off, to light up, the entire array fired up PERFECTLY!

This strange situation lasted for FOUR months!

That January, I get a very long distance phone call from my boss. (a genius, IMHO) He's on the scene because the lights are SUDDENLY acting up. (!)

Pulling on my memory cells - and gagging on the tale I'm hearing - I instruct him to the magic, offending box, the ONLY box that had any splicing whatsoever. I hate to splice on a 'dedicated' home run -- but these were truly long. (These two lighting circuits were the only runs in the 3/4" EMT, period.)

It's at this time, my field superintendent informs me of my man's boner.

Which left BOTH of us stumped: how in the world did the system run for me?

&&&&&&&

I had to conclude that the weather had changed.

Further, that the lights ran back in August because the extreme length of the neutrals acted -- for a while -- as a 'sink' for the harmonics. This effect was so pronounced because of the size of the returns -- #8 -- that they were able to bleed off their harmonics into the inductance of the very long EMT run.

Wetter air changed this equation enough to cause the electronics to shut down, entirely. (High quality electronic drivers will shut themselves down when they sense that something is really goofy with the wave form/ current differential.

As for the electronics, they HAD been able to draw out enough energy to pump the fixtures until then.

Why?

They actually constitute micro Secondarily Derived Services for each fixture. Each, in turn, was capturing energy and transforming it into a 'bucket' of DC -- in a series of capacitors. (A higher power version of the switching logic inside a DMM.)

It's from this source that the solid state logic builds a totally new wave form.

In sum, electronic 'ballasts' can tolerate pretty filthy AC power. Beyond some set limit, their logic shuts off any attempt.

This is fundamentally different than analog systems -- IE real world systems. For them, the gear tries -- and FRIES.

{ As an aside: you wouldn't believe the grief I had getting the crew to wire these lights. It was one of the first critical efforts that I didn't have the time to scissor up and inspect every box. I had to re-cap my boots after that marathon.

I'd SWEAR that the crews were gaming me: I was treated as a shuttle cock between 'crisis' after 'crisis.'}




Last edited by Tesla; 02/26/15 11:40 AM.

Tesla