First off the building is a county shop thats built in 1998.There has been two other Electrical company doing repairs on the shop. There was a fall out with them and the county. SO now im the one fixing all the problems. I have one major problem with the out side pot lights. There is no prints for the building so im learning as i go. Everything is three phase 208/120V. Out of the 32 pots only 4 work.They are the proper voltage 120v checked. The others are 255V!! I have no clue or never seen this before. The other contractors just replaced the par38's. Then the next day they burnt out. What or why is this happening?Im lost I looked and look and everything is spliced so so. Help

Is it possible that you're checking the voltage at the lampholders, and that they were/are metal halide lamps?

They are available in PAR38

It sounds like a either an open neutral within the circuit or a neutral and light leg crossed. You mentioned that everything is 120/208 three phase or Are you talking about the lights in questioned?

Voltage at the lamp holders. They and just a normal par38 light builb standard.First think I check was if they were MH with a regular buld dropped in.I found no balast just direct connection.The lights never worked since new. And there is no way to check how everything was wired. Main panel 208/120 lights are off of B phase. Hot 255V neutral vols.008 to .8. The same electrical contractor wired 21 175w MH around the shop. They installed the wire behind the bulb and I found 3 lights that burned and the rest have crispy cooked wire. The breakers never tripped pannel is FP stablock. There is also a ton more problems ill show pix latter on.

Not sure I understand. Hot 255V neutral vols.008 to .8. ...What?
The fixtures never worked and the building is from 1998 with FPE Stablok? Sounds like the electrical contractor was cleaning out his basement.

Have you tried using 130V PAR 38 bulbs?

Just thought,...with 32 fixtures, I’m guessing you would probably have several circuits to deal with. If they are fed with multiwire branch circuits and you have a loose neutral connection somewhere when under load, then you could potentially turn a lot of those PAR 38’s into flash bulbs.
You might start by checking connections in all of the light junction boxes themselves.

Everything is feed from four breakers. FP means Federal Pioneer 15amp stablok's. When I put the meter from ground to hot im getting 255volts. I tryed the same with the neutral to see what the voltage was. Sometimes I explane not to well. Im more of a talk in person guy than a typing guy.Spring they are istalling new sofets So i need to for now temp fix it. I thought maybe thay have a par 38 250V to temp fix for now. Ive pulled open all the pots and everything seems ok. Unless there is a broke wire somewhere.

Sounds like the lights may be fed with a multi-wire circuit and the neutural is broken or disconnected.

high leg?

The last 2 responces sound right.
Have you checked for ballasts?
Your description is vague.
Take a deep breath,Step back, and we can probably walk you thru it.

Your description is vague.Sorry not a big typer.

No Balasts just a normal pot light check all of them.
I think the neutural is broken but where. I checked all the splice and they look good. Unless there is a juction box somewhere with a bad splice. The bad thing is the building has been botched bad.

RDK, Based on what you have told us thus far, Here is what I suggest. First, fix what has burned up and disconnect the light circuit(s) in questioned. Varify that power and system is kosher. The burnt equipment could be the source of the problem.

Once that is all done. Varify you have the proper voltage on each leg going out to the lights.

Then ask yourself, "If I was going to run the conduits for the lights, how will I do it?" This can give you a starting point where to start looking. Once you can confirm conduits and wire routing, document it as an as-built. Check the connections. At each piece of equipment, check to see each light is rated for the voltage and that you have the correct voltage at each light. Document your findings.

I suggest if you feel comfortable doing this is if possible, locate a junction in the wiring that is in the middle of the circuit. De-energize the circuit(s), properly open the circuit(s) after properly marking them and protect all exposed ends. Energize the circuits. This will tell you several things. Whatever is energized and de-energized will tell you what side of the opening in the circuit(s) each load is on. If the goofy voltage dissappears, the problem is on the de-energized portion of the lighting system. If the voltage is still goofy, the problem is on the energized portion. Keep on repeating this until the problem is found. It may not be easy but highly effective. The theory here is you can have a ciruit with 1 million good splices and 1 bad splice. You can find that bad splice within 20 guesses by using this process. Keep in the back of your mind that there can be multiple problems and they can be on both sides in the open circuit so do not get tunnel vision. Also do not let burnt out bulbs trip you up.

When tracing circuits and conduits, do not be suprised to come to a relization that there is a junction some where like a hand box or junction box that you are not aware of. Keep us posted. I like these type of problematic issues. I know, I'm a sick individual.

The bad thing is ive done this all with no luck. PLus the whole building is mostly lumex.Ive started at the middle and worked from there. Seems like the problem is before the first light on all four circuit(s). The Burnt lights are on another set of breakers in the shop. The pots are in the office side from a diffrent pannel.

This almost sounds like you have 2 different systems crossed... Do you have 480Y/277V in the building at all? I'm not sure how you can get 255V otherwise, especially if all is well in the panel (aside from being stablok ) Even if this were a high leg system, you'd at best have around 2o8V-G. If no ballasts are involved and your panel voltage is 208V phase to phase, you're picking up voltage from a source outside of that panel to get 255V.

Thats what im thinking. With no prints or as builts im sorta lost. No 480/277 check all pannels.

Can you find which circuit breakers in which panel are feeding these lighting circuits?
I would recommend investing in a circuit tracer and toner/tracer set if you don’t already own one. Especially since this building sounds like it’s going to be "your baby" from now on. They can sometimes make problems like this a lot easier to tackle.

Since we are shooting in the dark anyway, is it possible these lights are hooked up to an autotransformer dimmer or something that could be incorrectly wired and somehow acting like a boost transformer?
If you are sure it’s a 208V supply, even with a bad neutral connection, I can’t see how there can be 255V anywhere in the building without some type of transformer ahead of it. I just noticed that you’re in Canada though and I’m really not at all familiar with the type of materials and wiring practice used there .

A bad neutral couldn't get 255V from a 120/208V system. At most it could be 208V
As a matter of fact, the situation you're describing is impossible electrically without some other influence.

255 Volts is a reading typical of a metal halide ballast's open circuit voltage.
Have you taken a mirror and a flashlight and looked through the hole for a ballast possibly mounted remotely? Some of them can be a real nightmare and take a real long arm to replace.
Taken a L-N voltage reading at the fixture's junction box line terminals instead of the lampholder? With a solenoid type tester?

I have to second electures above comments and those of Lostazhell, even with an delta service (or open delta) "high leg" there would be no way to get 255 volts as line voltage, an open neutral does not explain it.

There has to be a ballast/transformer/step-up influence of some kind at work here. The four that work are normal incandescent par 38 lamps? What is the voltage at thier lamp socket? what do the fixture nameplates if any say? I need more info to attempt a diagnoses but find the 255 volt report revealing in that there simply has to be a step-up process of some type involved. Let us know what you find....

In Canada it is common to use 347 volt for lighting which comes from the 600 volt three phase.

Now that I think of it, you are right about the open neutral The floating voltage will not exceed the phase to phase voltage.

The problem about trouble shooting on a message board is like a doctor treating a patient over the phone. You get tunnel vision.

I agree I did not see the part about the 208v supply the first time I read the post, tunnel vision I guess

jdevlin he specifically said in his first post that the service was strictly 120/208volt 3 phase. Also a 600volt service would still not explain the 255volts seen at the fixtures. He is missing some element here and consequently so are we, hopefully he will let us know what it is.....

Found the problem. The last guy who got fired installed a transformer up in the attic. Ill try to post some pix latter this week. THis was the same guy who installed a pannel over top of someone's crapper. Ill also post pix of that. Thanks to all. Now all the lights work

RDK,
So then, the 28 lights were Metal Halide fixtures with a single, large VA, multi-circuit transformer?
Can’t say that I have ever seen a transformer like that before.

I see a broken neutral. Start at the service FROM the meter IN the panel and measure the volts from each side to ground or the EGC - not the neutral. If it's anything but 120 volts - high, or low for that matter, the neutral is broken somewhere before the panel, - maybe the meter perhaps? Then measure the voltage in the panel from neutral to the ground or EGC connection. Voltage from neutral to ground of anything over about 2 volts indicates a busted neutral somewhere. Hope that helps you.