SOLVED(4 separate problems!) Help with tripping GFCI and washing machine

[MyEEVBlogAccount]

EDIT:  So far this seems to be fixed; see my later post for what is hopefully a permanent solution.  Thanks for the help and suggestions.

Before we begin, I want to mention that I am not an EE or electrician.  We are doing renovations on part of our house, of which the electrical work is done and, I've been told, inspected.  Ever since part of the "old work" was disconnected (which consisted of 5 regular outlets), the GFCI outlet that our washing machine was attached to started tripping during the same points in the cycle, with around a 25% probability.  The trips always occur a split second after the top-loading washing machine stops filling due to the solenoid closing at either the initial fill or the rinse fill, but not when it is added for any other reason (e.g. permanent press). 

Subsequent to that development, the breaker that the circuit that the GFCI outlet is on was moved to another subpanel, as the old one ran out of room.  Ever since then, it trips twice every cycle at the same points (e.g. when the solenoid shuts off after a long fill for initial and rinse).  I've tried using an extension cord to try it on the other GFCI on the same circuit (they are in parallel and it normally has our primary sump pump and battery backup connected), which tripped itself AND the other one on the fifth test, but none of the prior ones.  I've also tested it with the "new work" with an extension cord, but neither the new GFCI outlets nor the "AB breaker" (which I was told made the whole electrical compliant) it is connected to tripped during those tests.

The general contractor is less then cooperative with the idea of checking it or calling out the electrician, for various reasons, and we are hesitant to force the issue.  I've considered escalating it the issue, but if it was something simple to fix myself, I'd rather do that.  Does anyone have any ideas of what I could try?  Could it be the wrong kind of GFCI, if there are kinds?  I've verified the outlet is wired correctly with a tester and, as mentioned, it worked before with the additional parallel outlets without problem.  I've attached a basic diagram.  Thank you for any help you can offer.

[Richard Crowley]

I would simply replace the misbehaving GFCI with a new one.  They aren't that expensive and it can't hurt.  They DO age and go bad, after all.  Or at the very least, swap the misbehaving unit with one of the new ones and perform some "differential diagnosis" to narrow down the cause.

[MyEEVBlogAccount]

Yeah, I was thinking about just buying a new one, but the old one is only 9 months old or so.  It just seemed odd that the problem cropped up and got worse as changes were made.  But then again, coincidences do happen.

Would any from https://www.menards.com/main/electrical/light-switches-dimmers-outlets/electrical-outlets/c-9526.htm?queryType=allItems&Spec_OutletType_facet=GFCI&sortby=priceAsc work or do you need something special?

IIRC, the one currently there is this one https://www.menards.com/main/electrical/light-switches-dimmers-outlets/electrical-outlets/electrical/light-switches-dimmers-outlets/electrical-outlets/smart-electrician-reg-15-amp-self-test-gfci-outlet/9241-15iv/electrical/light-switches-dimmers-outlets/electrical-outlets/smart-electrician-reg-15-amp-self-test-gfci-outlet/9241-15wh/p-1490250024908.htm

[soldar]

The clothes washer might be OK (for certain values of OK) and the GFCI might also be OK (for certain values of OK) but the combination of both might be marginal and it could be that swapping the GFCI might resolve the problem.

The easiest thing is to begin by swapping that GFCI with another one and see what happens and what that may tell us. With that the problem may remain with the washer, it may go with the GFCI or it may disappear. Do this personally only if you have the knowledge and training to do it safely and without burning the house down or electrocuting yourself or others in the vicinity.

I have some wonky experiences with GFCIs. Sometimes problems need a little detective work to diagnose correctly.

Most appliances have some level of parasitic capacitive coupling between the live wires and the Protective Earth. It is usually low enough to not cause problems but if you have several appliances in parallel that small fault current can add up. Suppose you have a 30 mA GFCI and have 20 mA of parasitic losses. Now you only have a 10 mA margin. Suppose you have a GFCI which is specially sensitive and triggers at 25 mA. We are running near the limit and a small transient can trigger it.

A motor or solenoid valve might have too high capacitive coupling with PE and depending on what moment of the cycle it is connected it could have an inrush peak enough the trigger the GFCI. Computer PSUs have capacitors to PE besides whatever other unintended losses they may have. Individually they may not be much but add up a whole bunch of appliances in a household and it may add up to near the GFCIs limit.

Let me tell you a couple of stories. Note that the code in Spain requires Whole house GFCIs so only one GFCI for the whole house. This is a bad idea to begin with. Also note that we use Schucko plugs which are not polarized. Another bad idea. The problems I have encountered here would be greatly diminished or non existent in America.

Some years ago a friend told me his home GFCI was tripping and, after some observation, he had determined it was always in the early afternoon when he was working on his computer and would lose a ton of unsaved work. After several hours spread over many days my diagnosis was that the apartment had several appliances with parasitic capacitive leaks and the general GFCI was running near its limit. One little spike and it would be triggered. The refrigerator, the clothes washer, the dish washer, the computer, all had ever so small so small individual losses which added up. This problem would not happen with several distributed GFCIs. Further investigation showed it was the dishwasher which was doing the final triggering as it did the dishes after lunch. Reversing the dishwasher's plug seemed to resolve the problem but I recommended he install two GFCIs in the general panel, one for the kitchen (where most of the losses were) and another one for the rest of the apartment.

Another case, also some years ago, was that my own GFCI was also tripping and it seemed to be the clothes washer. Again, to make a long story short, the main cause was not the clothes washer but a toaster oven which had way to high parasitic losses and the clothes washer just added the tiny spike needed to trip the GFCI. The toaster over had (and still has) a resistance coil inside a metal tube and the capacitance between the resistance and the grounded tube was enough to have some serious capacitive loss which pre-loaded the GFCI. I further insulated the metal tube from the frame and housing with some tubing and that resolved the problem.

In my apartment in Madrid I have the GFCI protecting kitchen and bathroom but not bedroom and living room. I consider that safer for my computer data and even though it does not meet the Code I consider it acceptable as there are no women, children or animals here.

In my other house, which is much bigger, the GFCI protects the whole house and I have had no issues so far. Still, I have an induction cooktop which has relatively high leakage and, since I hardly use it, I keep it unplugged.

Leaks to Protective Earth can sometimes be a bitch to diagnose.

[Calambres]

I live in a three story house, being the two upper the "living" ones and the lower is a big garage where I have my "shop". I have one CFGI (automatic reset) for the two upper and another (manual reset) for the garage. I've had my own share of problems with CFGIs, which I reported  in this topic, and one of the conclusions I extracted from my research is that the nominal mA of the CFGIs is not real but approximate, being always lower than indicated. If you sum up this to the facts Soldar just related in the previous post about summing parasitic capacitances in some of my electric appliances, my house is always on the verge of CFGI tripping. I'm also planning to add another CFGI to minimize the risks.

In the cited thread I showed my DIY CFGI tester.

[soldar]

More GFCI anecdotes.

Some months ago my neighbor asks for my expert help. He tells me they went to sleep and everything was OK. He woke up to pee in the middle of the night and there was no power. The GFCI had tripped in the middle of the night. He reset it and everything was normal. His question was what tripped the GFCI? How could I know? It could be anything in the entire house. A good argument for separate GFCI on each circuit. At least you know what circuit it was and what is on that circuit.

A couple years ago the general GFCI in my house tripped a few times randomly. I had no clue as to where to start looking. I finally discovered the cause by chance and when I wasn't looking for it. A split type air conditioner dripped a drop now and then on a CRT monitor just under it and that caused the GFCI to trip. I noticed the occasional drop and fixed the aircon drain but only much later did I notice the GFCI had stopped tripping and connected both things in my mind.

I saw a swimming pool pump that tripped the GFCI. I do not think it was "faulty" as in degraded but rather that it was just built in such way that it had some serious capacitive leak to Earth. I guess it could be faulty, I don't know. The owner said it was too expensive to replace and he would just run it without the GFCI. I advised him this was dangerous (it's a swimming pool!) and he said he would only run it at night.  Later the police arrested some kids who had jumped the fence in the middle of the night to go take a dip in someone's pool. The kids mentioned feeling a certain tingling when in the pool. I think the pool owner was lucky nobody was seriously electrocuted.

I read that in America it was considered to require full house GFCIs and some houses did it but they were worse than what they were trying to prevent because old ladies would light a candle and venture down to the basement to reset the GFCI and would trip down the stairs in the dark and break a leg and/or start a fire. Finally it was settled that GFCIs at the outlet were better.

[Ian.M]

I would always be suspicious of any electrical appliance that uses (or removes) water.  The odds of it developing electrical leakage to ground are significantly higher than for 'dry' appliances, with the possible exception of ovens and microwaves that can experience condensation from cooking steam.

Unfortunately stateside, PAT testing is not commonly done, but you do need an *INDEPENDENT* electrician (with the appropriate certification, test equipment and training) to come and test all the appliances that are usually used on the affected circuits, and if there is no clear 'smoking gun' , carry on to test the electrical installation.

Yes, it will cost an not insignificant amount of money, but IMHO its worth it to determine whether or not you've got a defective appliance that could kill someone if the GFCI fails to trip, and if not, to get enough evidence to be able to force the contractor to fix whatever problems there are on the electrical installation.

As others have mentioned, a small capacitive leakage to ground is normal with modern appliances.  Its usually dues to the filtering required to keep the electrical noise from the SMPSUs commonly used in modern devices and appliances, from escaping up the supply wires where it could interfere with the normal operation of other appliances.   IIRC up to 0.75mA ground leakage current is permissible for most devices, and USA GFCIs with a nominal 'must trip' current rating of 5mA may trip at as low as 3.5mA of leakage current, so running as few as five devices or appliances in safe good condition, from a brand new GFCI could cause it to trip even though neither the appliances nor the GFCI have any fault.

[Gyro]

The usual reasons for earth leakage trips on washing machines (in the UK anyway) are (1) heating element internal leakage current and (2) Carbon deposits on brushed motors (combined with a bit of humidity). It's usually the heating element though, normally an easy replacement.

[soldar]

I have not ever measured it but I think water heating elements inherently have high capacity to the earthed shield pipe because they have a resistance wire separated from the outside tube by a thin insulating tube. That's quite a capacitor right there and if the insulation degrades over time then the situation gets worse. Electric motors I suspect also have quite some capacitance between the windings and the grounded parts. These days they put more insulation than there used to be in the past but, still, I suspect some significant capacitance is present.

That is why these are not Yes/No problems but problems which require understanding and evaluation. If you have a washer or oven with relatively high leakage then you can diminish nuisance GFCI trips by unplugging those appliances when not in use.

[Ian.M]

Time to do some measurements then!

However if you calculate the area of a parallel plate capacitor with an alumina dielectric of only 1/32" thickness, required to get 0.75mA current when 120V 60Hz is applied across it, you'll find its about 241 sq. in. which is a very large heating element, with very thin insulation.

[Seekonk]

I got a free LG washer with the common stripped motor rotor. Still an easy machine to fix. When I opened it up I was surprised to also see the water valve assembly was leaking. This is extremely rare.  Could be some actual leakage due to water spraying on the solenoid connector since you mention filling.

[johnkenyon]

Get a sensitive clamp ammeter. Clamp all wires feeding the appliance EXCEPT the earth wire, turn on the appliance. (For US 220v appliances that means the three 110-0-110 wires, otherwise it just means the two hot/live and neutral wires)
The ammeter will show any current which is flowing into the appliance, and out down the earth wire - it is this current flow that triggers the GFCI/RCD.

As it's been previously noted, you can get nuisance tripping if there's more than one appliance/outlet being fed by the GFCI and the sum of the current leaks to earth is greater than the trip.

Personally my money is on the heating element - washing machine fills with water, shuts off solenoid, turns on heat - trip...

[Gregg]

VERY important: is the neutral and ground only connected in one place? --- That should be at the main breaker panel.  If a sub panel has a bond between ground and neutral, it could be your whole problem.
Since you stated that the GFCI doesn’t trip on all of the cycles (like permanent press); assuming it uses a mechanical timer, I would start by taking the timer out and cleaning all of the contacts.  I would also clean off the fill solenoid valves where the wiring connects.  It is likely that lint dust has gathered on various surfaces and attracted moisture and old pitted contacts are not breaking quickly. 
Are the hot and cold water pipes connected at the water heater [and to the gas line if it is a gas heater]? It could be that you are getting some ground leakage via the water from alternate sources that are adding to the problem.  As previously stated, these leakages are additive.
Most US washing machines have a separate ground connection, usually a sheetmetal screw on the back panel, for attaching a separate ground wire.  It wouldn’t cost much to try adding a wire.

[HackedFridgeMagnet]

If you want to sleep at night, just get someone in who actually knows what they are doing and has the tools to diagnose then fix the problem.
They might find a fault in your house wiring that your washing machine problems are just a symptom of. Such as a second Earth Neutral bond like Gregg suggested.

Look under E for electrician in your local directory.

[sambonator]

...even though it does not meet the Code I consider it acceptable as there are no women, children or animals here...

Are you saying men are more resistant to electricity than women?   

[Gregg]

...even though it does not meet the Code I consider it acceptable as there are no women, children or animals here...

Are you saying men are more resistant to electricity than women?   

I doubt men have more resistance; hard headedness maybe.  Big Clive, for example: https://www.eevblog.com/forum/chat/big-clive-video-10ma-hand-to-hand-electric-shocks/ 

[Electro Detective]

You do not just swap out a GFCI (RCD Downunder) without FIRST externally testing that circuit with a decent GFCI tester, 
because the TEST button on the suspect GFCI won't tell the real story. Been there... 

You might have loose terminals and or moisture in the machine or an insect remains parked across the teminals, or in the power point

Everything from the machine to the power outlet and switch, may need some TLC before laying blame on the GFCI and mains board

Sometimes a GFCI (RCD) just plays up due to age, weather battering,
or you scored bad luck with an annoying intermittent oversensitive POS that needs tossing in the bin,
but only after pulling it apart for inspection if you can be bothered

Any decent electrical contractor should be able to track down the cause/s using generic AC-DC low and high current leakage testers and the usual methods.

Sparky 'mates' are to be avoided in a situation like this, unlikely they'll truly suss it out
and a waste of beer money 

[Gregg]

A lot of general contractors are licensed to install residential electrical but many do not have any real understanding of possible problems or newer technology.  Also some will use the apprentice that can’t use a measuring tape to install the wiring and not properly check on their work.  General contractors in the US seem to have a reputation of not quite finishing jobs.  Was the electrical upgrade permitted and inspected by whatever authority issued the permit?

If OP can find evidence of improper wiring, it would be reason enough to call the contractor back and demand a “proper” electrician certify any new work with the threat to report the incidents to the contractor licensing authorities or hiring an independent electrician to certify it at the general contractor’s expense. 

Many neutral to ground bonds are difficult to detect because a lot of cheaper breaker panels use a #10 machine screw that looks like it is supposed to be there as a bonding connection.  A picture or two of the sub panel would help.  Also if both neutral and ground wires are landed on the same bus in a sub panel is a good indication that whoever installed it didn’t know what they were doing.

[Electro Detective]

There are lots of electrical license wielding wankers out there, in a hurry to 'finish the job' to hurry on to the next halfassed 'install'

Many I've seen perform their sloppy work as if it's a paling fence knockup or grass n weed cut  (I would doubt their skill at that too  )

Their apprentices can be hit and miss at best, especially when their wanker boss is pushing them to 'hurry up' pulling/forcing through cables and teminating quickly without pausing to check   

They waste more time correcting their blunders and doing comebacks, rather than getting it right the first time, and save on stress and agro.

Some blokes do it proper in all respects, via current AS Rules/Standards, good materials and install expertise
and don't try and rip off the customer with 'electricians prices',

or if they have work on already, they over-quote... just in in case they get lucky and score an unsuspecting cashed up wood duck.   

[MyEEVBlogAccount]

Well, I had an essay written here, but then I accidentally hit the back button.  So this version should be shorter.

I plan to replace the GFCI with one of those from the link above as that seems the simplest thing to do at this point, although I'm not sure about brands but I think they should be about the same.  I also plan to recheck all the exposed connections that the GC touched when he disconnected the 5 outlets. 

I also wanted to make clear that the washing machine trips at the same 2 points in every cycle, and it doesn't matter what cycle is used or if the heater is on.  It also doesn't trip when water is added to the cycle at other points, like the permanent press spray during spin.  And once you reset it, it works just fine until the next fill is over, and always right at the end.

Additionally, the only thing downstream of the bad GFCI is the washing machine and an unused plain outlet on the "load" connection, as my GFCI three-light tester trips that outlet when I press the button in that outlet.

As to who did what, the GC disconnected the 5 outlets, everything else "old" was not touched by anyone since renovations started.  The licensed electrician did everything else electrical, including adding the subpanel, adding the new work breaker, and moving the old work one.  My spouse was here during the inspection, without the electrician, and the inspector made the GC tell the electrician to change the breaker for the "new" work to an "AB" breaker (as in the initialism, it is also marked with an "AB" and a test button).  All the other breakers are regular ones, as far as I can tell.  FWIW, the inspector did not take the cover off the panel, according to my spouse, so whatever that implies about the inspector / contractor / electrician.  Maybe that means it wasn't a "final" inspection, if there is such a thing?

Finally, I think we're getting close to the point where if a new GFCI doesn't fix this problem, then we will talk to the city inspection office.  We may not be able to force the issue, but the inspector might.  But that might also be a long shot, so I've debated getting the washing machine out and seeing if there is anything obviously wrong with it, like a leak or loose wire or corrosion.

[mjkuwp]

I think there are different kinds of GFCI. 

https://www.ul.com/wp-content/uploads/2014/04/ul_GroundFaultProtectiveDevices.pdf

in my house (Wisconsin, USA) we only have GFCI outlets near kitchen sink and in the bathrooms - so any place near water.  The Washing machine is not one one to my knowledge but now you have me curious to try putting one on there and see if things keep running.

[Gregg]

Personally I would not buy the cheapest outlet of any kind, especially one with anything electronic.  (I cringe at all of the ones available with USB power supplies built in.)  I would suggest a brand name like Leviton, Hubbell or Pass&Seymore; there may be others.  A number of big box stores hire semi-retired professionals for their electrical departments, if that is the case where you are, ask the old pro for a recommendation.  A few pennies saved are not worth the future risk.

[TheNewLab]

It may be the initial current surge in each cycle. the motor kicking on. But GFCI and breakers are supposed to allow for such brief current surges. A bad GFCI? possible has already been mentioned. OR a cheap one? ditto.
You might want to note if the GFCI trips when the cycles first start of later. If it does you have learned something..I would not hesitate to approach the city inspectors office, if you do not it get worked out. Always safe than sorry and any good city office would appreciate a concerned homeowner.

[MyEEVBlogAccount]

Before giving the solutions to the 4 separate problems discovered, I want to answer mjkuwp.  I was told it had to be a GFCI because it is 1) located in a basement in the floodplain, and 2) it is located a foot and a half from a utility sink.

So, ready for the story.  Over the weekend, my spouse did an unbalanced load of laundry.  It tripped at the start like usual, but despite making a hell of a banging racket made it through the rest of the cycle without the second trip.  Instantly piqued my curiosity.  So we are down there and I am looking at it while the second load is started and she says that the machine sounds funny, so we look carefully and the basket has moved inside due to the shaking.  She notes that the water is spraying in from a slightly different place, and going straight into the basket instead of hitting the side and sort of splattering.  Seems like just enough water is splashing to trip the GFCI, as it makes it through the next load non-stop.  (Problem 1).  Unless we either do a larger load or use the (NEMA 10-30) electric dryer, the washing machine continues right past the usual tripping points.  But still tripped whenever the larger loads are done or the dryer is turned on.   So close and yet so far.

So, the electrical inspector, who is also a licensed electrician, comes out today.  Apparently, the general contractor has had problems with accurately describing the scope of work, following code, and "seems to think he's a sparky."  He looks at the electrical.  While inspecting one of the connections the GC did, which he had assumed was good since the electrician used was on their "good" list and wasn't on the "new work" list anyway, the neutral conductor comes out of the wire nut while is is maneuvering it out of the box, which means it probably wasn't even in the nut and caused a bad connection.  (Problem 2).  He also discovered that the water pipes in the home were not properly grounded when the city installed the new water meter, resulting in a leakage.  (Problem 3).  Finally, looking at the dryer, he discovered a leak "from the dryer neutral from the timer tap" to the metal shell (he explained that 3 plug dryers don't have their own ground) which was then leaking through the ground of the washing machine as the two were physically touching across the sides (Problem 4).  He thinks the combined leakages were enough to trip the GFCI, and also said that repeated tripping can cause the GFCI to become more sensitive and suggested replacing it regardless.

Hopefully all this means that the problem is taken care of once and for all when I replace that given all the other things that were fixed.

[Alex Eisenhut]

GFCIs can also be wired in series to protect a regular outlet somewhere else, is there anything else plugged on that circuit?