Is This a Floating Neutral ? What's Going On?

I’m curious, what does this look like to you all? A bad neutral? It started right as a kitchen renovation started, right as they started the demolition. The demolition didn’t get anywhere near the breaker panel.

It kept up. Sense only lets me post 5 at a time :upside_down_face:

I should note that around the 21st is when I noticed it. I did not get a notification from sense, but it’s when I heard the very quiet alarm from my Siemens FS140 ground reference monitoring feature. The problems are still ongoing according to sense, but the alarm hasn’t sounded for several days. It didn’t seem to go away after switching breakers off, either.

@test35965 ,
What are you seeing from the “Floating Neutral” part of the Sense Labs ?

ps: I think I’m going to start this up as a new topic.

1 Like

This is a floating neutral. The neutral should be balancing the voltage.

1 Like

This is not from the “Floating Neutral” detection in Sense.

It’s the kind of thing that would be useful for sense to detect!

Now I’m curious, why do y’all think it happened when the demolition started? Do you think it’s at the breaker or at an outlet? This is a 1983ish home. I’m not sure if I should check individual outlets, since I’ve heard about olden days use of shared neutrals, and many of these damn outlets were backstabbed. They were drunk when they built the house - I found beer cans from the 80s buried in the attic insulation :joy::man_facepalming:

1 Like

Now check this out. On the morning of the 28th, I tied some of the grounds together. The grounds I tied together are in outlets that were removed on like the 22nd pending the rest of the renovation. We left the house on that day, and also sense is a few days behind on this feature, so it’s too early to say what’s going on.


A couple thoughts.
First, my view is that that the Sense “Floating Neutral” detector should have caught this. I’m wondering if the the Labs “Floating Neutral” detector is experiencing some of the same operational issues and/or delays in being run that have occasionally beset the Power Quality checks. Since this is in the Labs area it is not guaranteed to be functional all the time, but I’m wondering if @JuliaAtSense or someone from the support team can check in on this. BTW - here’s how the Floating Neutral detector is supposed to work (yours is a no-brainer case in my mind):

As for what could have caused this to happen - I think you only get this phenomenon (voltage legs going in opposite directions) from a floating neutral at the breaker box or at the utility transformer, not from a broken neutral on a circuit (which can also be dangerous). Did the remodel any work inside your breaker box / distribution panel ? Or did the utility make any changes ? Sometimes the issue is just things like corrosion.

You can test for an open neutral on a specific outlet with one of these outlet testers.

1 Like

It appears to me that you do have a problem with your neutral cable.

Areas to check:

  1. Visibly check the ground connection from your breaker box to either the ground rod located outside your house or if it is connected to a metal plumbing pipe inside your home. Take pictures and post. You could have corrosion at the outside clamp. Or possibly the connection to your plumbing pipe ground was disturbed and is no longer connected. Either way, power everything off and clean up the connections with a wire brush. I had to clean and install a new ground clamp at my home. After installing, I got better voltage readings.

  2. Check the outside wires coming to your home. If you have an overhead line entering from a power pole, try to visually inspect the bare neutral ground wire. My brother had squirrels chew thru the aluminum bare neutral wire. It wasn’t completely eaten thru so his voltage readings were similar to yours. The power company had to completely replace his service drop to his house. It was free. If you have a ground transformer, have the power company check their equipment. Show them your graphs. It’s possible that their ground rod located at the transformer is not connected or has failed. Take pictures.

  3. If you have an external electrical shut-off and can access the inside of the breaker panel safely, then remove the panel door and check for a loose or ‘burned’ connection on the white/neutral cable. Also check the green ground wire. Take pictures. You need to check the main neutral connection point in the panel. The lugs might need to be tighten (to 25 ft-lbs). Wire is soft and will deform over a period of time. If you live in an area with saltwater corrosion, you should consider using an anti-corrosion grease on the wire and lugs (like NOALOX). If you don’t feel comfortable doing this yourself, then you should definitely contact a licensed electrician to provide a service check. The cost of hiring someone will save you money in less failed electrical appliances.

  4. Your ground wire from the breaker panel to the ground rod could be too small.

Keep us posted on your progress. Good luck!


Here’s the funny thing, yes, they did add a new circuit, but only several days after the first disruptions recorded by sense. Btw, sense here has been invaluable in diagnosing this. If it weren’t for sense, there might have been some nasty finger pointing. My best guess is that somehow the vibration and general movement from the demolition jiggled something loose somewhere.

1 Like

To my frustration, I don’t know where the exterior ground is. I will go on a safari when I’m back at the location! It’s a condo, and the three units were attached to have their meters ganged next to eachother. I do know that the power lines are not overhead, that’s clear to me :wink:

#3 is not super hard for me insofar as inspection is concerned. I’ve had that panel cover off a bunch of times and have a healthy respect for the dangers of electricity. I’m comfortable enough with live power to work around it, but not stupid enough to work on it. I hire pros when actual work needs to be done to the breaker panel.

A few years ago, I had it off to air seal the panel… If you’ve never checked before, you’d be shocked how much infiltration can come through panels if there’s no caulk in the penetrations to the attic or the building wall spaces.

Anyways, I did actually take the panel off so I could look at it with my little FLIR attachment for the camera for exactly this reason. Torque screwdrivers weren’t even around when this place was built, so I figured there’s a chance something might be loose and causing a high resistance connection. I didn’t see anything. Hence the confusion. Even more confusing, the ground reference monitor alarm was intermittent. I could actually tell because the Corsi Rosenthal boxes in the house would slow down intermittently. The intermittent part is what made this super frustrating when it first started. Nothing seemed to help!



I just wanted to check in here and say that I saw this thread. I can confirm that what @test35965 experienced was clearly a floating natural, and I’m glad it was caught. With that, the question goes, why didn’t it appear in the floating neutral section of the labs’ app?

What I can say is I’m working with some folks to figure out why. What @kevin1 already said is true; this is a lab’s feature and may not work all the time; it is also possible that based on the particular situation Sense didn’t pick it up.

This is helpful info to hear so we can work to improve and get us all looking a little harder at what to expect from floating neutral detection.



I experienced one floating neutral, 40 years ago, but they are memorable events. Mine was due the connections in the meter socket - my round meter had stabs that plugged into a socket. If my memory is correct, the underlying cause was corrosion between the copper stabs and aluminum lugs, but it has been 40 years. You can google that for better explanation.

From that experience, and as an EE, I suggest you focus on the neutral connection (the big bare wire at the service entrance, which ties to your neutral bar with all the white wires). Poor grounding can be a contributing (secondary) issue, but is not likely the root cause. When you lose your neutral connection, the two 120v legs, what we call the split phases, become a simple series circuit across 240v. You have lost the primary reference back to the center tap of the distribution transformer. Your ground wire can act as a secondary path for the neutral reference, but its resistance back to the transformer will vary and add to the confusion.

Without a solid neutral, the voltages on each leg float based on what is called a voltage divider principle. Perhaps when you turn lights on in Room A, you notice the lights actually get brighter in Room B. (My exact experience). That is because as you add loads on the Room A side, you actually give the electricity more paths to flow, and the parallel resistance goes down on that set of 120v legs. Due the voltage divider principle, the voltage on the Room B side will increase, such that the sum of Room A and Room B voltages always equal 240v. This phenomena will be present on all circuits common to the Room A side and all circuits common to the Room B leg

I cannot offer an explanation why this occurred when demolition started. Floating neutrals should always be the result of something amiss at the service entrance. I don’t want to think about more complicated scenarios where the house neutrals wired together somehow compensate and provide a neutral, that would be a sign to me of poor work and I just dont have enough understanding to run that one through my brain.

Hope this helps, others please chime in if you see something amiss in my explanation.

Terry B


Ok, this is getting pretty frustrating! I had my electrician replace the failing main breaker outdoors at the service entrance. We thought that would fix the problem, never mind that it’s good we replaced a clearly bad breaker. I was kinda annoyed, because he’s very busy and decided to do it while I wasn’t around, so I couldn’t get him to use a torque wrench. Electricians never torque things unless you’re there, asking them to do it, standing there watching them, and also usually loaning them your own torque wrenches (:triumph::rofl::face_with_symbols_over_mouth:). I don’t even know if he abraded the strands first, they usually don’t do that either.

However, the problem persists and is getting worse! Sense picked up high frequency noise indicative of arcing (which, again, I love sense so goddamn much). It’s extremely clear there’s a floating neutral at this point. The problem is that it’s intermittent. It’s mostly fine most of the time. But then it’ll act up, and at one point I had 160v on one leg for a whole minute. Another point, about 2 weeks ago, it went full line to line a few points, and blew out the power supply on my CPAP. Through the surge protection, which probably all needs to be trashed now.

I had the utility come out. Of course, we couldn’t trigger the fault when he was there. It’s like debugging really terrible code! We tried everything. I tried putting 14kw of load on the house. I tried putting several kw of inductive only loads on the house. He used the fancy toaster to test the busbars. He checked the padmount transformer. His only suggestion was to keep calling whenever I spot the problem.

We have power through a 4 position MeterPak enclosure, which has a lot of aluminum components, and an aluminum 4/0 service entry cable. I’ll be very frustrated if this is an aluminum corrosion problem, and probably write some letters to UL and various NEC committees asking to speed up the adoption of copper coated aluminum instead of friggin aluminum. If people are going to try and save money by not using 2/0 or 4/0 copper, they should at least be forced to use something with better long term oxidation safety!

We have an old QO load center as the main indoor panel. In the past I could swear I heard arcing in the panel, but can never ever see it happen with the cover off.

The lugs are reasonably tight. You’re not supposed to check torque by applying listed torque values to a bolt that was tightened 40 years ago, so I screwed around by shutting off the main breaker outside, set my torque wrench to a very low value - like 25% of the listed 250 in/lb torque - and checked those main lugs. Nope. Tight.

Where the copper grounds are lazily twisted together, and passed under a screw down lug with the 4/0 aluminum service entry, which is a shit way to do it, I set my wrench to the lowest setting it would register (probably around 60ish inch pounds) and torqued that. I got about two turns out of that.

The breakers themselves are the plug on type. No screw retention. If Sandy Munro has taught me anything from his excellent YouTube videos recounting his long and ongoing automotive engineering career, it’s that threaded fasteners should be inferior to just about anything else when it comes to reliability. And yet, these breakers feel a bit wobbly. Several weeks ago I dug through piles of square d docs and they do seem to suggest with confidence that the plug on design wipes corrosion off the busbars and also retains well. They explicitly say to not use grease of any sort… But it looks like there’s some grease on the contacts in my panel. There’s no evidence of heat damage. Some of the connections seem discolored, but not any more than I have seen anywhere else in this house, given none of the electrical boxes were air sealed, and constantly exposed to infiltration. Yes, I’ve been going through a lot of DeoxIT when swapping outlets :rofl:

Not gonna lie, I’m very frustrated and also grasping at straws. The other day I was doing some non-electrical work in an electric box (caulking, because contractors still think “a house needs to breathe” and so they make no attempt to airseal boxes, let alone but the airfoil airtight boxes) and got a shock on my left hand between the neutral and ground wires during a sudden voltage fluctuation.

Does anybody have tips or suggestions?

Areas to check:

  1. Visibly check the ground connection from your breaker box to either the ground rod located outside your house or if it is connected to a metal plumbing pipe inside your home. Take pictures and post. You could have corrosion at the outside clamp. Or possibly the connection to your plumbing pipe ground was disturbed and is no longer connected. Either way, power everything off and clean up the connections with a wire brush. I had to clean and install a new ground clamp at my home. After installing, I got better voltage readings.

  2. Your ground wire from the breaker panel to the ground rod could be too small.

NEC 2020 code now requires 2 ground rods placed more than 6-feet apart. Buy another ground rod and some additional clamps. Placing a second ground rod may not solve the problem, but the suggestion is cheap and easy to implement. Do you live in a dry desert area?

1 Like

Futzing about with the grounds is on this week’s agenda.

The service ground is definitely not undersized, it’s 4/0, but it is aluminum, and it’s connected hilariously farkakte. Yeah, all the home grounds are twisted, and then everything is shoved under a lug or two, with some noalox vaguely smeared around. Not the most reliable connection.

Now that I’ve decided that I no longer trust the panel either, I am going to shut the power to the house and try to rebound it to best practices. McMaster-Carr is always a savior, and I grabbed two parts that will help:

7915K14 Multi-Way Tap/Splice with Through Hole for Two 6-4/0 Gauge Wires

6920K53 Tin-Plated Aluminum Set Screw Lug, 1 Hole for 2 Wires, 6-250 MCM Wire Gauge

At this point I honestly don’t think it will make a difference, but it’s worth a shot to try one of them.

The set screw lug is shinier and feels way more skookum, but the insulated tap with through holes might also be useful. The 2 hole lug should be able to fit in where there currently is the one hole lug. I think the torque for both of these will be pretty darn high… UL recommends 375 inch pounds for 3/8 hex set screws!

1 Like

Would you please post additional pictures:

  1. Show the whole breaker panel.
  2. Take a picture of where your grounding wire actually connects to your ground rod.

A new ground rod only costs $17.98 at Lowes or Home Depot
ERICO 5/8 in. x 8 ft. Copper Grounding Bar Rod 615880UPC - The Home Depot

NEC 2023 electrical code now requires 2 ground rods for residential service. Drive the 2nd ground rod 6-ft or greater from the 1st ground rod and connect both ground rods. You can use an acorn clamp - $2.98/ea.
Southwire 5/8 in. Grounding Rod or 1/2 in. Rebar Ground Rod Clamp for #10 SOL/STR - #2 STR Wire 65176440 - The Home Depot

Do you live in an area that has dry soil or damp soil?

You are correct. Torque specs for wire connections are high. It always feels to me that I am about to break the lug connection to attain the correct torque. After I initially tighten a lug, I will come back a day later and recheck by re-tightening again.

Sure. Here’s a picture of the whole breaker panel:

Here’s the ground rod:

Unfortunately I cannot modify the connection to the ground rod, as that’s shared equipment with other homeowners, and the property manager does not believe there’s an issue with any shared equipment. He refuses to do anything, especially anything that costs any money, unless someone can definitively locate the problem somewhere in their equipment. Which makes this a lot more annoying.

Just in case, I did remake that terrible connection in our inside panel with the large McMaster Carr al/cu rated lug. Abraded each strand of aluminum while it was coated in penetrox, got it under the lug, and then all the copper wires under their own lugs. Torqued to 375 in/lb.

There’s also a ground bond of SOME sort near the water heater. Don’t know where it goes, but I took it all apart, blasted it clean with DeoxIT, shined up the pipe with some abrasive, then put a smidge of penetrox at the spots where the clamp connects to the pipe and where the copper strands go under the screw for the zinc clamp. Torqued the connections down to like 45 in/lb or so (according to the UL suggested torque values for screws of that size).

Here’s a curious thing I’ve learned over the past few months that I don’t think I mentioned:
The problem is much much worse when it’s rained for several days, and almost entirely absent when it’s dry out. Even a bit of rain doesn’t do it.

The power company came out last week and installed a high end power quality analyzer to the busbars in the MeterPak to see if it’s a problem on their end. It will be there for two weeks. Of course, we’ve had little rain, so we’ve had no problems. I might have to stand about 20 feet away and make some “rain” with my hose for a while. Can’t afford to wait until the weather changes :triumph:

It’s a NEMA 3R panel, it’s not gonna kill me to stand far away and rain water down, but also, this is why I will stand far back.

Since you are not the owner of the ground rod, then that presents a problem. Plumbing grounds which were acceptable, can now be problematic. You never know if someone has replaced a portion of the copper plumbing with PVC or PEX piping and now that ground is lost.

The only other option that I can think of is to purchase a spray can of CRC 2-26 electrical spray and saturate the outside grounding bond. Maybe that will penetrate between the metals and help with the connection. It appears to me that your outside grounding connection is corroded.

My ground rod pictures:

Main ground rod at the house -

Secondary ground rod located near the garage (more than 6-ft from the main ground) -

New residential installs require one continuous (uncut) ground wire. Since I could not replace the whole ground wire back to my breaker panel, I ran a second connection using two clamps. I took a wire brush and cleaned my outside connections. I was only without power for 5 minutes. It didn’t cost me any money to clean the connections. You do need to power off all electrical to do this!

I also have copper plumbing. My cold water inlet line to the water heater is ‘bonded’ to the hot water outlet. This is to slow galvanic corrosion. Is your plumbing bond for house-hold electrical or plumbing corrosion control?

You posted some good pictures! If you are allowed to make changes to your breaker panel, then I would suggest that you add a whole house surge protector inside your panel. That would protect your large appliances and add a secondary level of protection for computers and TVs.

After looking more closely at your breaker panel picture the problem may be more with the neutral connection than the ground connection. The inside panel is not bonded, so the neutral-ground bond must be in the outside main shut-off. It is possible that you have a loose neutral lug on the outside box. That might require the power company to re-tighten. Do you have pictures of your outside electrical? Can you open it up?

Good luck and keep us posted on how you solve this problem.

Indeed I do have a whole house surge protector! If you read up at the top, it’s quite a nice one. A Siemens FS140. The neat thing is its Ground Reference Monitoring feature properly alerted me to this problem before sense did! It is supposed to clamp around 150v or 160v, and did so for many months, but I recently had to replace (before fixing this underlying problem) it because it no longer could do this, and the excessive voltages damaged a few things. Highly recommend this unit.

1 Like

The strange part of that ground is I’m not sure where it goes. It doesn’t seem to end up in the panel, so it must be for some other bonding purpose. The pipe it’s attached to, is definitely a low resistance ground. It goes directly into the slab. It runs probably under the slab.

I can’t disconnect that rod. There’s a sign on it that says “call Verizon before disconnecting!” Or something to that effect. Which is odd, because the ground wire definitely goes into the MeterPak, not the comms boxes on the same wall which supply an internet connection to the whole neighborhood and also across the highway!

Correct that the inside panel is not bonded. It’s really a “main lug loadcenter” not a “main panel”. There’s a 200 amp breaker outside that we already replaced. We actually replaced it because it seemed to be beat up and corroded.

The neutral cable does go out to the MeterPak, and connects under a lug behind the main breaker. No matter what I said, I couldn’t make any of the licensed electricians who’ve come out to remove the neutral cable from the lug, abrade it while coated in fresh penetrox, reinsert, and torque to specs. Which was irritating. This part of the system was our responsibility, we didn’t need the utility. In theory, I could do it, but I’m not an electrician and I also do not own the proper arc flash PPE… There’s no way in hell that I’m actually going to do it unless there’s a gun pointed at my head. That’s live 240v, exposed, and unprotected. Not something to mess with :rofl:

Everybody insists the connections are “tight”.

If there’s a problem somewhere downstream of the utility cables and also upstream of our main breaker, then it’s an HOA problem. Our electrician offered to disassemble the part of the MeterPak panel where we heard arcing when the power company was out, but there’s concern that disassembling it would damage the 40 year old panel, and we wouldn’t be able to get replacement parts. The HOA would then have to replace the whole panel, for several thousand dollars. So they don’t want to touch it.

1 Like