Big Voltage Drop Under Load

[Smart $]

so why does a bad neutral cause such a voltage drop

It does not cause a voltage drop in the typical sense, i.e. that which results from conductor resistance and long lengths.

What does happen is it turns your electrical system into a parallel series voltage divider. The neutral carries the imbalanced of current to maintain line to neutral voltage. When that flow is impeded or cut off, voltage is not maintained line to neutral while line to line voltage is still the same.

 

[gar]

080726-0916

fireryan was referencing my mis-statement:

If the GECs are wired correctly and there are no shorts to the GECs, then there is no current in the GECs. Even if there was a short to a GEC this has nothing to do with the unbalanced voltage problem. Note the GEC is bonded to the neutral at the main panel and thus only originates at the main panel.

Replace GEC with EGC to correct my paragraph. Problem is disconnect between thought and correct acronym, and maybe whatever point in another post on which I was commenting.

Looking back at post #29 I see I read it wrong.

The answer to post #29 is that yes there is a current thru the ground electrode with a lost neutral or high resistance neutral from the transformer. This I described in post #37.

.

 

[fireryan]

thanks for clarifying

 

[gar]

080726-1144 EST

fireryan:

The relationship of the resistance from the transformer center tap to the main panel neutral-egc-gec point to the load resistors on each phase determines the unbalanced load voltages.

I wrote an equation for one of the load voltages relative to the values of the three resistors of interest. This was a complicated equation and I did not want check its correctness. Nor would it give you an intitutive feel of the problem. Thus, I presented the example with a known current in the neutral leg because it was simple and illustrated the problem.

There are two rules to always keep in mind. On an instantaneous basis the sum of the voltages around any loop are zero, and the sum of any currents at a node are zero.

.

 

[mark32]

The other day I posted about getting a voltage reading at the panel of 107 at a co-workers house. Initially I thought there maybe a problem with the service drop but after talking with a contractor friend I believe it may have been from neighbors beings supplied from the same transformer. It was 5pm, everyone getting home and cranking their A/C's up, sounds logical right.

By the way, Gar, I enjoy your posts. You're a troubleshooting guru.

 

[gar]

080726-2110 EST

mark32:

Many air conditioners will be 240. Assuming no neutral problem, then you should see a low voltage on each side of neutral.

If there is a heavy 120 load on one side of neutral, then that voltage should be low and the other side might increase a very small amount.

What might be the source impedance of the transformer. Assume 25 KVA and 5%. The transformer current rating is 104 A. 5% of 120 V = 6 V. Impedance is 6/104 = 0.06 ohms. Add some service line drop, but that is local to your load, and the voltage may be lower due to your own load. How much current might be loading the transformer from several neighbors? How much current does a typical air conditioner draw? It is quite possible the power company has reduced voltage to shed some load. This does not appreciably reduce the air conditioner load, but it certainly does on incandescent lamps and stoves.

.

 

[cschmid]

red horse simple way to find the loose neutral get a infrared non contact thermometer and check the temp on all the connections this will get you in to the area..most likely you can get the correct connection point..this is only good for exposed connections..so you might have to remove a couple of covers..unless you own a infrared camera that is best..

 

[electricalperson]

red horse simple way to find the loose neutral get a infrared non contact thermometer and check the temp on all the connections this will get you in to the area..most likely you can get the correct connection point..this is only good for exposed connections..so you might have to remove a couple of covers..unless you own a infrared camera that is best..

good infared thermometer is the fluke 62. it does the job and its pretty cheap and small enough to throw in your tool box

 

[quogueelectric]

Why would that be? :-?

The current will always try to return to its source of emf. The problem is the xformer is grounded at the neutral and the main service disconnect is grounded so if you lose the neutral between the xformer and the main the groundrods try to complete the path through the dirt and they are not capable of carrying much current under a real life load. They are there more to carry lightning strikes and overvoltages safely to ground than to complete the path of a broken neutral wire. It may help to stablize the voltages on both sides to a degree but in no way is safe. As the current increases the voltage drop also increases quickly raising and lowering the voltages in the residence depending on the loads as they begin to ignite and or blowup and clear. The voltages just keep bouncing back and forth until the fire dept gets there and pulls the meter.

 

[LarryFine]

The current will always try to return to its source of emf. . . . etc.

I know all of this stuff. I thought it was being suggested that the EGC between the house main and the pool panel was seeing voltage drop.

The point I've neen trying to make is that the only way an EGC could see 'normal' current was if the poor neutral was at or ahead of the service.

Once the neutral and EGC become separate conductors, nothing should be able to induce a current on the EGC (unless driven into the system.)

 

[ramsy]

loose neutral...in the panel, the meter, weatherhead or on the poco side....send me a check for $165, that's our diagnostic fee....

Just did one yesterday, $160 for two hours. Erratic lights & appliances. Utility fingered outbuilding weatherheads, but won't touch sub-panel feeds. Started in rigid, then crossed buildings using #6 solid as both messenger and neutral.

Subpanel had all N-G under Blue wirenut, but it worked perfectly for 30 years. Same 135 - 99v imbalance w/out wirenut, no signs of arcing.

HO demanded it stays grandfathered as is, and there is no permit required, no 30yr old codebook, or authority to force new raceway, or ground rods for modern code compliance.

Got to play with feeders & weatherheads for a while. A Shuretest measured voltage drop an order of magnitude larger from L-N than L-L. Also, lifted feeders from breakers to the service ground, then found 30k Ohms L-N at subpanel, but Zero Ohms L-L.

So, I re-stripped the loose, 30yr old, corroded copper on each exposed weatherhead split bolts, and replaced w/blue wirenut & tape. Then found Zero Ohms L-N across the de-energized & temp-bonded feeders. If using IR camera, don't know how visible that voltage drop would have been on messengers blowing in the wind?