Floating neutral or wiring problem?

[scorpster]

Last night several of the lights and outlets went dark in the house. Then it mysteriously came back by itself for a few hours, and now these circuits went out again. All fuses tested ok (unless there is a failed circuit breaker somewhere). One of the dead circuits is an outlet that is right below the main panel (so it's not a fuse or breaker in a remote subpanel). Looking into this further, I tested the voltages going into a subpanel that is right next to the main panel.

The difference between the two hot phases going into this subpanel show

110v. Between one of the hot phases and neutral it is 120v. Between the other hot phase and neutral it is only 12v!

Last year the electric company came out to fix a broken neutral wire on the telephone pole, and this reminds me of the floating neutral problem I had back then, causing mysterious problems with some of the circuits. Because there is a major difference between the different phases and the neutral, I wonder if I should call out the electric company again, maybe their original repair failed?

Or could this be due to a wiring problem inside the house that I should check first? Or perhaps a failed circuit breaker in this subpanel?

[ransley]

Call the Electrc Co now , I would caution against using anything unless a V meter has pointed to show its ok, I had bad repairs on the electric co

[RBM]

If you test between the two hot legs, you should get 240 volts. If you test from each hot leg to neutral, you should get 120 volts. If you got 120 volts between the two hot legs, you have a dead leg. You need the check the breaker or fuses feeding that sub panel

>

[scorpster]

All the 15A circuit fuses in the main panel check out ok when I test-replace them with a new fuse. I get 0.2 Ohms on the large 100A, and 40A fuses so they're fine too. So now I'm thinking maybe it's not a floating neutral, but rather it could be a dead leg on the phone pole? Maybe one of the hot legs could have broken on the phone pole from the swaying tree branches just like the neutral line connector broke last year.. would that be consistent with these symptoms?

[RBM]

You want to test at the main breaker, preferably with the main turned off. You have three legs coming in, 2 hot, 1 neutral. With the main breaker in the off position you should get 240 volts across the 2 hot legs, and 120 volts from each hot leg to neutral. If you don't get 240 across the 2 hot legs, one of them is dead

[Don Young]

Your problem is not a loose or floating neutral like you had before. You have a bad connection or breaker on one of the main hot lines. I would not recommend that you try to repair this. Call an electrician or the power company.

Don Young

[scorpster]

Mystery solved! Southern CA Edison sent over a repair expert tonight within a couple hours. Turns out that one of the hot legs was broken at the telephone pole. The neighbor's tree branches are completely tangled around the wires, putting tension on the wires and causing the entire pole to tilt at an angle. This unraveled one of the hot wires. He repaired the connection, and he's sending a crew out within a week to trim the neighbor's trees, and then replace all the wires between the two poles, for a permanent fix.

The best diagnostic clue that I found before he arrived was to measure between the two 100A fuses at the main panel, where this showed 0v instead of 240v. The reason I still got 123v from between each phase and ground, he explained, was because of back-feeding from the house circuits.

Tip: keep your trees trimmed near the phone poles, and keep an eye on your neighbors trees!

[Tony Hwang]

Hmmm, Your mistake was using that Fluke with too an input impedance. One reason I still often use Simpson 260. Some times El Cheapo anlog meter is better in such as our case.

[The Daring Dufas]

Everyone should have a Wiggy. It needs no batteries and it's hard to break.

TDD

[scorpster]

Which readings would have most likely been different if I had used a better measuring device? I thought the Fluke multimeters were pretty good but it sounds like high voltage electrical is more tender.

[Tony Hwang]

Hi, Fluke is very good digital meter but it is TOO sensitive due to high input impedance(which means meter does not rob current from the measured circuit loading it down}. When you had funny reading, you could try a lamp to see if the reading was real or not. Digital meter has a typical input impedance in the 1,000,000 Ohm ranges. Analog meters have 5,000 to

20,000 Ohm ranges

[stan]

Exactly. One of the legs (sometimes but incorrectly called 'phases') was open.

The two legs, with 230 volts between them are actually the two ends of a single 230 volt transformer winding; on the pole or underground etc. The middle or centre point of that winding is the zero or neutral point. A neighbour had the identical problem with the 230/115 volt service to his garage. Broken connection at the pole transformer. He couldn't figure why his 115 volt lights worked but nothing 230 volt would work!

In non North American jurisdictions several actual 'phases' are sometimes brought into domestic/residences etc. In the Middle East my son had 230 volt 3 phase main circuit breaker panel in his 'villa'.

In Malta the three phases and neutral (4 wires) went along the street on wall brackets and a different phase (and the neutral) were tapped into each house!

And over in, for example, the UK everything is 230 volt; no 115 volt (except in certain special and unusual cases).

Here, in Canada three wires come into house one being the neutral. The other two being the 230/115 volt legs as described by the OP.

[stan]

[stan]

Even cheap (non-Fluke) volt meters are sensitive enough to pick up stray voltages, if you connect one side to ground and then touch your fingers to the input or + lead one will often pick up enough random electromagnetic radio and other electrical 'noise' to get small reading.

And connect it to even dead wiring such a meter will often pick voltage from other working wiring running next to it within walls etc. Posters here have reported 'stray/random' but meaningless voltages up to 43 or more volts.

A good choice for testing such a situation is a plain bulb. Or a pocket neon lamp/tester (Maybe that's what's called a wiggie?).

A 230 bulb is best but a low wattage 115 volt can be used if just touched on wires for a moment.

BTW I have a 230 volt low wattage bulb permanently mounted in my workshop which monitors the 230 volts; if one side were to go open the

115 volts from the oher ;side' would come through whatever 230 volt appliance happened to be on and that bulb would light more dimly than at 230 volts dimly. The single bulb is a better way to monitor the 230 volts than two bulbs one from each 'side or leg' to neutral.

If one needed to test best way is to turn off main breaker and see what voltage is on the incoming wires .............

OR: Turn off all the individual circuit house breakers (every single one of them) and see what voltage there is on the output of the main breaker by testing the buss bars or inputs to the individual breakers.

[Tony Hwang]

Hmmm, Pole? I don't see any pole in my neighborhood. Power, phone, cable, all are under ground. Nothing overhead. 4 wires, one is ground.

[trader4]

Capacitance is not an induced voltage. And there's no reason you can't use a digital VOM on AC. That's what I always use.

Rather than ignore them, what you should have is an understanding of basic electricity so that you can correctly interpret what you are reading. As others have said, you can get a voltage reading with a digital unit because it has a high input impedance compared to a typical analog unit. Which means that it disturbs whatever it's measuring a lot less. The analog VOM actually loads down the line it's measuring enough that the 8 volts becomes .8, or whatever. You can turn the digital VOM into a less precise instrument if you want to by putting about a 1K resistor across it's terminals, effectively lowering it's input impedance.

[HeyBub]

Not if the resistance was quite high.

If I = E/R where E=220 and R=(more zeros than the stimulus) the resulting I (current) is teeny.

If R = (say) 220K, you get I = 220 / 220,000 = 1 ma.

Power, then, is EI = 220 x .001 = .220 watts or 1/4 watt.

[Doug Miller]

1000-ohm resistor at 110 volts passes 110/1000 = 0.11 amps 0.11 amps at 110 volts = 12.1 watts

Not if it's only dissipating 12 watts.

[trader4]

Yes, you got me on that one guys. I didn't have my brain fully engaged. I was thinking more along the lines of the principle, which is you can turn a high impedance digital meter into a lower one by using a resistor in parallel. Something along the lines of 100K would work, which would give you 1/2 watt at 220V.

[Ralph Mowery]

Fluke makes excellent meters and test instruments. I have a couple at home and 3 differant types of testers at work. One is a very good modle 87 , one a simple digital volt and ohm meter with an amp probe, and one I like the best is really simple. It has several leds on it and tells if you have AC, Dc nad a very rough indication of the voltage,or low resistance in the circuit. I also use some of their high end (over $ 4000) calibration instruments. I deal with almost all kinds of industrial electricity. Everything from low voltage 4 to 20 miliamp signals to 480 volt 3 phase circuits at 500 amps or so. Also some 4160 volt circuits and while I have not done anything with it, we do have some 13,200 volt main feeders to deal with.

YOu just have to know how and when to use the meters. For most simple AC problems I use the old Simpson 260. If I suspect a false reading due to induced voltage I change the scale and if the meter stays in about the same place , it is induced voltage.