Voltage Drop Mystery

Users who are viewing this thread

BLH

New Member
Messages
4
Reaction score
0
Points
1
Location
Utah
This one has me stumped. I'm not a Pro but not a novice electrician either. Advice and suggestions appreciated.

Included on my home's master panel is a 100 amp breaker that feeds power to an external garage. Voltage on both leads to and through that breaker is about 124 volts. The wire connecting the two buildings is aluminum #2 AWG, or possibly larger. It's big. This home was originally built by an electrical contractor as his own home. He didn't take shortcuts. There are 3 wires including the ground wire. They run through an underground conduit. The distance is probably 50 yards.

I first noticed that my garage door would energize but only barely start to open, then stop. (I could get it to open under power if I manually assisted.) Next I noticed that half my florescent lights were flickering. I tried running a table saw and it would run but not at normal speed and it "sounded sick".

So Ok... I'm getting a power loss somewhere, but where?

I first suspected the 100 amp master panel breaker itself was bad, and indeed, it probably was, so I replaced it. Didn't solve my problem, however.

After replacing the breaker, I again checked the incoming power on the external garage's panel. Voltage on one line was about 120, as expected. Voltage on the other was about 90. That's with all breakers in that panel off and no load whatsoever. Putting a load on (like turning on the florescent lights) dropped the voltages to about 100 and 70 respectively. Putting more load on (turning on the table saw) made those drop to 80 and 30.

I suspected a grounding problem and checked and tightened all ground connections in both panels. All seemed good and tight, but I removed the main ground line coming from the master panel anyway, checked it for condition, and reinstalled it.

Nothing helped.

I began to suspect a wire problem, perhaps a partially broken or corroded line somewhere in the conduit underground. To test that, I reversed the two wires at the master panel's 100 amp breaker, then measured the incoming voltages again on the garage panel. I expected to see the voltage drops reverse to match the wire change, but they didn't. The readings were still as they were before. That would seem to eliminate a bad wire problem.

BTW, this upper garage panel also powers in the summertime an underground deep well pump without any problems. There has always been plenty of power for that -- no notice of any dimming or power drops when the pump cycles or when starting up power tools, such as a compressor.

At this point, I'm out of ideas. Anyone out there have some to share?
 

Stuff

Well-Known Member
Messages
1,214
Reaction score
128
Points
63
Location
Pennsylvania
You reversed the wires at the main panel's breaker and the voltage drops didn't change at the garage. All that does is eliminate the main panel breaker and feed. Mostly reinforces a wiring issue.

It sounds like there is a wire problem or there is something else in the middle that you don't know about. Check for splice boxes leaving the house inside and outside. Are you sure there isn't a disconnect on the outside of the garage? Maybe a junction box there.
 

drick

In the Trades
Messages
453
Reaction score
14
Points
18
You have 3 wires and not 4? So two hots plus a neutral and no ground then. Your neutral wire is bad. When the neutral wire fails 120V devices like your garage door opener will not work or only get partial voltage. Your 240 volt devices like your well pump will run great because they do not need the neutral.

Check your voltage between the two hots. Is it 240V? Is it always 240V? If yes its a bad neutral.
 
Last edited:

WorthFlorida

The wife is still training me.
Messages
4,638
Reaction score
698
Points
113
Location
Orlando, Florida
Aluminum wire insulation, if broken anywhere, or just a pin hole, aluminum will corrode fairly rapidly. Maybe the conduit filled with water. When you measured the voltage across L1 and L2 you're reading the two hots only. Read from L1 to the white neutral and L2 to neutral. even with a corroded neutral and no current flow (no load) a voltmeter only needs a milliamp or two to work. Does the remote panel have a ground rod? If the garage door and lights are on the same circuit check the wiring that a varmint hasn't chewed on the cable or a bad wire nut connection in a juncture box.

Don't rule out a bad motor on the door opener and the lights are on the same circuit.
 

wwhitney

Well-Known Member
Messages
5,058
Reaction score
1,267
Points
113
Location
Berkeley, CA
Some preliminary comments, very general as I'm not sure what you know:

It's important to distinguish between the neutral N (or grounded conductor) and the EGC G (or grounding conductor). N is a circuit conductor and will have current on it during normal operation of 120V loads. G is a bonding conductor and will have no current on it during normal operations. The job of G is carry enough fault current to flip the breaker if a line conductor L (hot, ungrounded) contacts a bonded metal part of the electrical system that should not be energized. So a given system of EGCs should have a single N-G bond (to complete the circuit in the event of a fault), at the main panel.

Your 3 wire feeder is L-N-L, likely with no G. That's not allowed for new feeders, but is OK for existing feeders between buildings as long as there is no other metallic path between the buildings. So if you ever run a coax line, or a telephone line, or a metallic water line to the garage, you have to upgrade the feeder to have an EGC. When the feeder has no G, you need an N-G bond at the garage panel to reestablish the EGC. While when the feeder does have a G, you must not have an N-G bond at the garage panel.

How do you know there is underground conduit? A direct burial cable will be sleeved with conduit for the riser portion at each end. If there is continuous metal conduit, that could be your EGC. [And if it's old enough, it could be compromised now.]

A 100A breaker is oversized for #2 Al conductors, max is 90A. Or maybe your conductors are #1. Hopefully you can read enough of the writing on the insulation to determine the size. Or else take a calipers to the deenergized conductor.

Included on my home's master panel is a 100 amp breaker that feeds power to an external garage. Voltage on both leads to and through that breaker is about 124 volts.
[. . .]
After replacing the breaker, I again checked the incoming power on the external garage's panel. Voltage on one line was about 120, as expected. Voltage on the other was about 90. That's with all breakers in that panel off and no load whatsoever.
That's almost impossible. If each there's no current flowing on the feeder, then the voltage measurements at each end of the feeder should be identical. You were measuring L-N at both ends? [If you were measuring L-G at the garage, and the garage's EGC is compromised, then your measurements at the garage were phantom.]

Cheers, Wayne
 

jadnashua

Retired Defense Industry Engineer xxx
Messages
32,660
Reaction score
1,144
Points
113
Location
New England
Make sure that the breakers and panel are rated for AL wire, and it may need the anticorrosion paste applied then be properly torqued. Aluminum rusts, but ALO2 is essentially clear, so you can't tell it easily. It also is about the same volume as the aluminum, so it doesn't show rust like iron does. (Iron when it rusts, ends up larger, breaking away from the solid stuff beneath, exposing it so it can then rust some more.) ALO2 is also an insulator, so the anticorrosion paste and properly torquing the connection(s) is important.

A modern digital DMM might have many megaohms per volt load it puts on the circuit, so it's essentially looking at open circuit values. If there's any resistance in any connection, it is likely much smaller, so you would not notice.

If the subpanel has its own main breaker, that could be bad as well.

If you disconnected the L1 and L2 leads at their source, connected them together then went to the other end and check resistance from one to the other, see what you get in ohms. Then, back at the other end, take one hot lead, connect it to the neutral for that cable after disconnecting that, adn go measure the resistance at the other end and repeat with the other hot lead. They all should read about the same and be a small number, close to zero. There are calculators you can use online that will tell you the resistance of the wire based on the gauge and length. Compare that to what you read.

FWIW, in a properly working electrical system there should NEVER be current on the ground lead. IOW, it should not affect the operation of things plugged in. Ground is there for safety, to provide an alternate path to trip the safety device whether that's a fuse that blows, or a circuit breaker that trips.
 

BLH

New Member
Messages
4
Reaction score
0
Points
1
Location
Utah
OK... Thanks so much for all your help. The responses gave me much to think about and check out.

I've learned some things, and think I have diagnosed the problem, but in case I'm missing something, I'm going to walk through the process and show my path.

First of all, a couple of clarifications and answers to questions.

In my original post, I misspoke. I said:
I suspected a grounding problem and checked and tightened all ground connections in both panels. All seemed good and tight, but I removed the main ground line coming from the master panel anyway, checked it for condition, and reinstalled it.
Where I said "ground", I meant "neutral". Oops.

The garage panel box is grounded to a copper grounding rod in the ground below the panel. Also, there is a metal conduit coming from the ground, which is also grounded to the box through a separate grounding connector as well as through the pipe connection itself. However, at some point below ground the conduit changes to a 4" PVC, ABS, or similar. I know that because I can see a portion of it where it passes by a driveway drain grate. There are three wires running in it, thus L1, L2, and a N, but no separate ground. There is a slight opening in the conduit resulting from settling and earth movement, so I can see the wires. I also suspect that may have been an entry point if a varmint has chewed on my neutral wire, which I suspect may be the cause of this whole problem, but we'll see.

There are no metal connections (water lines, etc.), between the two buildings.

I have cleaned and tightened all ground connections and neutral connections on the garage panel, as well as the L1 & L2 connections on both panels. The connections between the master panel and the garage panel are lacking dielectric grease, but I will fix that soon.

With that as background, here is what I've found. I took these voltage readings at the garage panel as follows, under three conditions:

1) All breakers off
2) A single pair of breakers on, which power lights and a circuit in the building that I have a table saw plugged into, but neither energized.
3) As #2, but with the table saw running. It table one, it ran very "poorly", and obviously not up to full speed or torque, with some interesting voltage readings (see red cells in table one). In table two, with an improvised neutral wire replacing the original, it ran normally. (I cobbled together a temporary neutral wire to replace the existing one using some #12 extension cords running above ground).


Voltage Readings.JPG


So, it appears the neutral wire is compromised, which means pulling 70 yards of replacement cable through the conduit. I hope that is easier to do than it sounds. Question: Is it necessary that this neutral wire be AL to match the L1 & L2 wires? Could it be appropriately sized copper instead?

If anyone has any hints on the best way to connect and pull wire, I'm listening. This will be a first for me.

If you have any suggestions, comments or observations, please let me know! Thanks again for all your help!
 
Last edited:

Reach4

Well-Known Member
Messages
34,555
Reaction score
3,499
Points
113
Location
IL
I have cleaned and tightened all ground connections and neutral connections on the garage panel, as well as the L1 & L2 connections on both panels. The connections between the master panel and the garage panel are lacking dielectric grease, but I will fix that soon.
Read up on Noalox and similar.
 

wwhitney

Well-Known Member
Messages
5,058
Reaction score
1,267
Points
113
Location
Berkeley, CA
The garage panel box is grounded to a copper grounding rod in the ground below the panel.
OK, if ground rod(s) are the only grounding electrode at the garage, then two are required. They need to be 6' apart, 8' long, and fully buried.

However, at some point below ground the conduit changes to a 4" PVC, ABS, or similar. I know that because I can see a portion of it where it passes by a driveway drain grate. There are three wires running in it, thus L1, L2, and a N, but no separate ground. There is a slight opening in the conduit resulting from settling and earth movement, so I can see the wires.
Well, you should not be able to see the feeder anywhere it is underground. If it is PVC conduit, it needs 18" of cover (soil on top of the conduit); if it is direct buried, it needs 24" of cover.

Also, 4" would be very large. What color is the pipe you can see? Does the feeder pass under the driveway? If the feeder is direct buried, it may be sleeved under the driveway.

Your load testing certainly shows a bad neutral. In general you don't pull out just one wire of several from a conduit, and pull in a replacement; rather you pull them all out and and then pull in the desired set. In which case you should pull in 4 wires including EGC and remove the N-G bond in the garage panel. For the shortish distance involved, the higher cost of copper may be worthwhile, since you are replacing a failed aluminum wire.

I think there is a good chance that you don't have a complete conduit system that you can reuse. In which case you would need to trench from the house to the garage, deep enough for 24" cover if you are direct burying; 18" cover if you use PVC conduit ; or 6" cover if you use rigid metal conduit (and are not passing under the driveway).

Cheers, Wayne

P.S. The L1-L2 voltage ought to be the same in both of your tables. In table 2, did you perhaps disconnect the house side of the underground neutral? There could be an underground L2-N fault, which would explain your data, if the underground neutral was disconnected in table 2.
 
  • Like
Reactions: BLH

Reach4

Well-Known Member
Messages
34,555
Reaction score
3,499
Points
113
Location
IL
P.S. The L1-L2 voltage ought to be the same in both of your tables. In table 2, did you perhaps disconnect the house side of the underground neutral?
I noticed that. I would think that the lower voltages (table one) were at a high utility peak use time, and the bigger L1-L2 would have been a lower use time.

Fix your wires. But also consider asking the utility to change the tap on the transformer to give you closer to nominal 240 volts. Mine usually runs about 256, and I am thinking about asking them to turn it down. My UPS on my computer trips when the voltage gets to about 130 on the 120. Use to trip at about 128, and I adjusted the trip level up. See https://www.rockymountainpower.net/.../pp-rmp/power-quality-standards/1C_2_1_PF.pdf figure 1.

As to whether you could pull a new wire, you might see if you could push a wiring snake. Maybe mark the distances on the snake first, so that if you get stuck, you would know where to dig. Presumably that would coincide with a conduit failure. I don't think getting stuck with an electrical is likely, but easy enough to be extra-careful.
 
Last edited:
  • Like
Reactions: BLH

drick

In the Trades
Messages
453
Reaction score
14
Points
18
So, it appears the neutral wire is compromised, which means pulling 70 yards of replacement cable through the conduit. I hope that is easier to do than it sounds. Question: Is it necessary that this neutral wire be AL to match the L1 & L2 wires? Could it be appropriately sized copper instead?

If anyone has any hints on the best way to connect and pull wire, I'm listening. This will be a first for me.

Its going to be a pain to pull wire 70 yards, but 4 inch conduit is a good size, so you have that going in your favor.

Step 1:
Pull out all the wire, measure how long it IS so you'll know what length wire you'll need to replace it with. Throw the old wire it in the trash. 70 yards of wire that has been sitting in a conduit for years is not going to want to come out. You already have a break in the conduit at the driveway. I would cut away about 4 feet of conduit WITHOUT damaging the wires too much and pull in both directions from there instead of pulling from an end. You may need to invite a friend to lend a hand.

Step 2:
You will need the following: A Shop vac (bigger is better), duct tape, lightweight nylon string and heavier weight nylon rope the length of your conduit plus a little extra, two 4 inch repair couplings, and enough 4 inch pipe to replace what you cut out. You'll be needing one of those cheap lightweight plastic bags from the grocery store checkout so be sure to have one of those (or similar).

Next you're going to suck the lightweight nylon cord through with the shop vac. Decide which end you are going to going to suck from. If your property is level it doesn't matter, but if the conduit has any slope to it let gravity help you out, suck downhill. (So ideally your positioning the shop vac at one end of the break in the pipe by the driveway and feeding the string from either the garage or house.) Using the duct tape, connect the vacuum hose to the conduit so that you have a good seal between the two. Take the plastic bag and cut into something like three or four strips maybe 5 inches long by two inches wide. Lay them in a star pattern and tie them to the end of the nylon string. (Basically the goal here is to give the vacuum something to suck through the pipe. You want to fill the pipe diameter with the shredded bag, but not create too much drag that the vacuum can't suck it through. There is some trial and error involved.) Turn on the vacuum, insert your plastic bag on the string and if you're lucky the vacuum will suck the string through first try. Once the string reaches the vac, run the string through the repair coupling, the replacement pipe, the second repair coupling, and then repeat the process with the second half of the pipe. **If you can't suck it through sometimes blowing it through with the vacuum or a leaf blower works too.

Step 3:
Install the replacement pipe section, but DO NOT bury it yet. Next tie the rope onto the string and pull it through the entire length of conduit. (Make sure the rope and the string cannot come apart, otherwise you will be repeating step 2.)

Step 4:
Buy your wire. I would use copper, but its up to you. Make sure the wire rated for wet locations. I would upsize the wire because of the distance to reduce the voltage drop. Make sure the maximum ampere rating of the wire does not exceed the rating of your circuit breaker. You'll need four wires this time: red, black, white, and green. Buy enough wire. By that I mean at least 10 feet more that what you think you'll need.

Step 5:
Pull in your new wire. You do not want the wires separating from the rope so take the time to make sure it will stay together with 2 guys pulling hard on the other end of it. There are probably better ways, but what I do is strip 8 inches of insulation off one of the conductors, make a loop, and braid the two sides 0f the loop together leaving enough of a hole to tie the rope onto it leaving about 18 inches of rope tail behind the knot. Then I would place the remaining 3 conductors behind the first, staggering each one further back by a couple inches. (This makes the wire bend easier going through a turn). Using the rope tail I tie multiple hitch knots, one every few inches, to add some extra grip while pulling. I then wrap it in a good layer of duct tape to help hold it together and to keep it form hanging up on anything. Now all you have to do is pull the wire through the conduit. Having a person on each end helps. Good Luck!

So hopefully your pull was a success. If it was you can go ahead and bury the replacement pipe. If it wasn't a success and you were able to pull as far as the break in the pipe you could install a pull box at this location. Basically you would install two 90 degree elbows, come up above the ground 18 inches, and install a weatherproof box big enough to allow you to bend the wire 180 degrees without exceeding the wires maximum bend radius. I'm thinking with a 4 inch pipe you won't need to do this, but you'll have the option as long as you have enough extra wire to come up and make the bend.
 

wwhitney

Well-Known Member
Messages
5,058
Reaction score
1,267
Points
113
Location
Berkeley, CA
How about:

Step 4.5 Pull a mandrel through the conduit to clean out any debris and prove the conduit is clear

Is that a good idea in this situation?

[I'm going to be impressed if there is actually 4" electrical conduit buried at the proper depth, continuous between panels.]

Cheers, Wayne
 

BLH

New Member
Messages
4
Reaction score
0
Points
1
Location
Utah
EPILOGUE:

Thought I would "return and report" to resolve the mystery, and thank you all for your advice and help.

I dug and pulled and pulled and dug (nothing is every as easy as we hope it will be), and extracted the 200+ feet of #2 AWG wire. Thank goodness for a good friend who came to my rescue with the right equipment or I don't know what I would have done.

Anyway, at first I found that the wire had been spliced -- just below the master panel on the house, so it was obvious that someone underestimated when they bought the wire. I could almost hear their swear words!:(
Splices 2.jpg



At first I thought that was the problem, as splices are supposed to be known trouble sites, but the splices looked clean with no corrosion. Hmmm...

Splices 1.jpg


As we pulled the remainder of the wire, the problem spot was obvious.

Corroded Wires.jpg


We ended up trenching and burying another 70 feet of 3" conduit (to straighten the line and put it where it didn't go under cement). This time, I think it's been done right.

There is a euphoric feeling when all is done and you flip the switch and everything works! :)

Well, I learned some things from this and am grateful for your comments that gave me ideas and helped confirm my analysis. Thank you for your willingness to share your time and expertise to help a stranger!
 

wwhitney

Well-Known Member
Messages
5,058
Reaction score
1,267
Points
113
Location
Berkeley, CA
Glad it worked out.

Did you end up replacing with copper, or stick with aluminum? Adding an EGC, and making sure there was no ground/neutral bond in the garage? Two ground rods, bonded to the EGC in the garage panel?

Cheers, Wayne
 

BLH

New Member
Messages
4
Reaction score
0
Points
1
Location
Utah
Did you end up replacing with copper, or stick with aluminum? Adding an EGC, and making sure there was no ground/neutral bond in the garage? Two ground rods, bonded to the EGC in the garage panel?

I decided to go with aluminum, the same as was previously used, #2 AWG URD (3 wire). No ground or other connections between the buildings, and there were two grounding rods in place already, with an N-G bond in the garage panel. On my earlier post, I just hadn't looked close enough to notice both of them. Thanks again!
 

wwhitney

Well-Known Member
Messages
5,058
Reaction score
1,267
Points
113
Location
Berkeley, CA
I decided to go with aluminum, the same as was previously used, #2 AWG URD (3 wire). No ground or other connections between the buildings
The NEC exception allowing the omission of the EGC is only for existing feeders, so you were supposed to pull in a 4-wire feeder. Of course, the 3-wire feeder will still work like it used to.

Hopefully your URD is also labeled XHHW or RHW, or both panels are on the outside of their respective buildings. URD not so dual-labeled isn't allowed inside buildings, as it hasn't passed the necessary flame spread tests. So its use would require that the URD terminate (in a panel or a splice) on the outside of each building.

Cheers, Wayne
 
Top
Hey, wait a minute.

This is awkward, but...

It looks like you're using an ad blocker. We get it, but (1) terrylove.com can't live without ads, and (2) ad blockers can cause issues with videos and comments. If you'd like to support the site, please allow ads.

If any particular ad is your REASON for blocking ads, please let us know. We might be able to do something about it. Thanks.
I've Disabled AdBlock    No Thanks