So, #10 is the washer in the laundry room and #12 is other receptacles in the laundry room? Are there any other lights or receptacles on either of these two breakers?

Mtvhike - There is a ceiling light in the room, but it isn’t connected to the same circuit as the washer. I’m not sure whether it’s connected to the circuit for the other outlet.

Well, you could turn off breaker #12 and see if the light goes out! What else is that circuit used for? Your washer is in #10 and your dryer appears to be on #16 and #18.

If you're comfortable with changing breakers, I would replace breaker #10 with a non-GFCI 20A breaker and the washer receptacle with a dedicated GFCI receptacle. One thing would be nice is a single outlet GFCI receptacle (all the ones I've seen are duplex), but perhaps they don't exist.

Your quote of the post-2011 code says "Now GFCI protection is required for any 125-volt, single-phase, 15- or 20-ampere receptacle installed within the laundry area, even if no sink is present" seems to say that EVERY receptacle in the laundry area (room?) must be GFCI-protected. I wonder if that's actually true!

Talked to Leviton today to ask if they have heard of this issue and was told yes. Then asked if they have a washer only GFCI outlet to help with this issue and was told no do to the fact that they have guide lines that must be followed. Need to stay between 4-6 ma. Hope somebody can come up with an answer to this.

Robert Gann

Hey bob, I wish I knew what instrument was used to capture the data, but I borrowed it from the "Power Electronic News" website and should have gave them credit.

It's not from a washing machine but from a frequency inverter.

https://www.powerelectronicsnews.com/counteracting-high-leakage-currents/

This article has a solution toward the end the author said is simple but I'm not an expert and it will take me some time to completely understand what he's talking about.

illinois524

I would expect that response from Leviton. There is really only one set of GFCI specifications - and it does require a trip at 4-6 (5 nominal) milli-amps.

riellebee

It does look like your washer is on a dedicated 20 amp circuit per installation manual. This is not code, but it is required by most washers. The reason is that the washer is a big enough load that if you put other things (like irons ;0) on the circuit, it will likely overload and pop.

The single plug is typical for this type of circuit - it kind of "enforces" the "dedicated part" although not really - cause there could be other plugs downstream. However, it is typical for this type of install to put a dedicated 20 amp round plug and a GFCI breaker. From what I can see, your install looks right up with expectations.

As mtvhike says - if this were my house I would do what he says. Replace the breaker with a standard 20 amp breaker and put a receptacle type GFCI in for the washer.

Important notes:

- You should put in a 20 amp GFCI in this application. Most GFCI's are 15 amp - so you have to specifically find a 20 amp one. Be sure you find 20 amp with 20 amp feed through, not 15 amp with 20 amp feed through. The easy way to tell is that the plug will have one T-shaped hole.

15 amp will have just vertical holes. See the photos attached. The Ivory colored is 15 amp, the white is 20 amp.

- You really do need to know what you are doing to change the breaker. I can't recommend that to a "typical" homeowner. Once you pull the panel - you are exposing yourself to high voltage/high current on open terminals. Your breaker box has no Main - so you cannot deenergize the panel here. You would need to find the main disconnect. (An electrician would probably do this live, I would, but I don't recommend you do it).

That said, I do laugh at

"but the electrician said that the entire breaker would have to be swapped out."

That is like - really easy. The model number is covered by the panel, but your breaker looks like a Square D Type HOM Combination AFCI/GFCI. The standard breaker replacement for this is a Square D Type HOM 20 amp breaker. Costs $4.10 and they have 68 available at my local Lowes. But, you need to know what you are doing.

I've done it dozens of times - an electrician has done 1000's.

I cannot recommend it for a typical 'homeowner'. It is easy, but mistakes are lethal.

20 amp GFCI

15 amp GFCI

There are 20 amp single plug GFCI's, but they are hard to find. You don't really need one (the attached photo is of a Leviton GFPL2-PLW

mtvhike

You ask,

"Your quote of the post-2011 code says "Now GFCI protection is required for any 125-volt, single-phase, 15- or 20-ampere receptacle installed within the laundry area, even if no sink is present" seems to say that EVERY receptacle in the laundry area (room?) must be GFCI-protected. I wonder if that's actually true!"

Yes - that is my interpretation of the code. Every outlet in the "Laundry area" must be GFCI protected. The only thing that is not deterministic is what a "Laundry area" is. Usually that will be left up to the inspector.

kevin9408

Nice find on that article. The power inverters being discussed are just the type of variable frequency motor controllers we have been discussing.

"Frequency inverters in particular, which are needed for energy efficient operation of motors, cause large residual currents."

Now - this article is really about industrial installations - but the concepts apply. It is also clearly not US/NA because it is talking about 50 hz, and the limit they discuss is 30mA, which is European (US uses 5mA) but that really doesn't matter. It is also really for system designers/engineers - people who can put filters in their device. In our case, we can't so we fall to the approach described as:

"If multiple inverters are used in a system, it can be worthwhile to use a central filter at the grid input instead of a filter for each individual inverter. This not only saves money and space but also reduces the leakage current. Many manufacturers also offer special low leakage current filters for their inverters or summation filters for use at the grid input."

The filter I listed above is the type of filter they are talking about.

https://www.digikey.com/en/products/detail/schurter-inc/5500-2646-01/6005845?s=N4IgTCBcDaIKxwAyIHRgGwBZ0sQRhAF0BfIA

They also mention various levels of RCD (a Residual Current Device). These are two terms for the same thing

"A ground fault circuit interrupter (GFCI), or Residual Current Device (RCD) is a type of circuit breaker which shuts off electric power when it senses an imbalance between the outgoing and incoming current."

I've never seen similar ratings for GFCI.

HU-349741951

Ah Ha! Notice the European limit for RCD's is 30 mA. US GFCI's are 5mA. The Meile washer is German. Is it possible they are testing/filtering for 30mA, not 5mA?

Bob

It is not obvious from the data sheet at the link at Digikey that the filter will have enough attenuation at the frequencies shown in the spectrum analyzer output. However, using a filter is certainly an approach that in principle I would gravitate toward to cut my time losses dealing with techs and distributors.

So if a seemingly suitable filter is identified, it could be purchased as an experiment. If the test results are sufficient (no more breaker actuations), such a filter could be put into the washer (probably a pain to do), or into an electrical surface box with a cover that will take a receptacle suitable for the washer plug (mind the cubic inch ratings and requirements) with a cord and male plug with fitting in a 1/2 inch knock-out in the box. An electrician should be able to assemble such a box if necessary.

Should the customer have to do this -- no.

Should the customer have to pay for this addition and effectively significantly increase the cost of the washing machine -- no.

Should the customer have the right to keep trying washing machines until one is delivered that isn't problematic -- yes.

Should the customer consider what his time is worth -- yes.

In the USA, NFPA 70 requires Class A GFCI in many locations, including laundry rooms. These are 5 mA type. Where the GFCI function is intended to protect machinery, rather than people, 30 mA breakers are available.

There seem to be many countries where 30 mA is used for person protection. This is a current on the verge of initiating ventricular fibrillation, so might be considered risky.

I'm unclear how UL certification occurs when the device needs a 30 mA leakage threshold to operate but isn't going to get it in the USA.

Yes - that is correct. Europe and Australia as well as some Asian countries use 30mA, not 5mA. The Shurter paper refers exclusively to 30 mA. They also discuss Class A up to B+ which have different frequency responses. They seem to say, in that paper, that a Class A (which I think we are talking about) should be sensitive to sinusoidal frequencies only - which is what you want for this case.

kaseki

If you go to the Schurter page you can see the frequency response diagrams much more clearly.

Here is the attenuation curve for type N at 16 amps. You are correct, it is unclear that it is attenuating much at 10Khz (.01 MHz). I misread the scale on the small diagrams as KHz, not MHz. With that scale, the range of interest was 1-10 and there is lots of attenuation there. But looking at this correctly, we are down at the very left end of the scale, which may have 10db attenuation. Not sure that is enough......

... particularly as the attenuation is falling as it heads to 6 kHz and below.

I wonder what frequency that 6 mA spike is all the way to the left -- 60 Hz or 120 Hz.

There must be a filter out there that looks like the circuit of the example but scaled for lower frequencies.

Lower frequencies often mean larger components - which is unfortunate. It's also kinda hard to do these things as digital filters in high power.... I'll keep poking, but have not found a suitable filter yet. Using one of these might be a fun experiment, but....

This is why I earlier suggested SurgeX. Not inexpensive, but big enough packaging. However, I wouldn't buy one for this application without talking to them. (My SQ washer doesn't need one, and the breaker is a Siemens/Murray GFCI -- not sure at the moment from what year's batch.)

I've been looking at the SurgeX stuff. The problem(s) I have are 2 fold:

- No detailed electrical characteristics are available except via a call.

- These devices are designed to protect products in your house/office from an noisy supply (utility).

The case we are studying here is the opposite - the product putting noise on the line. It is not clear to me if the products would help in that way (maybe if you install them 'backwards') It all depends on the circuitry.

I've sent them an email asking that question. I'll post a reply if I get one.

- Bob

"I'm unclear how UL certification occurs when the device needs a 30 mA leakage threshold to operate but isn't going to get it in the USA."

At this point in time it has a lot to do with big money by it's not the UL.

Kaseki, UL is just a third party private certification lab, OSHA approved and certified by ANSI. The ANSI is a private non-profit organization which oversees most codes we live by including the national electric code (NEC), so It's not the UL but the ANSI That set the standards. But these are merely codes written by a private non profit company which ARE NOT binding by law unless a state, county or city adopts the codes as law, and they all do.

Safety is a good thing but some code changes make no sense to us but means big bucks to the ANSI. It needs to generate revenue and does it through the sale of code publication material, and membership fees from 270,000 companies. If nothing changes there is no need to buy the new code books, and no need to be a fee paying member (who in the back room may or may not lobby the ANFI to change a code so they can sell their products.) ANSI can only continue to exist if there is change, they don't get public funding so the finger is pointed at the ANSI. I may not live long after writing this but the ANSI and the membership companies with influence make Billions changing the codes.

UL doesn't actually test much product in their labs. It's all done by 4th party labs certified by UL to test items to meet ANFI standards by means of UL testing standards. 4th party labs are required do the tests the way the UL standards are written and it's kind of like a franchise in a way, We don't know what conditions the lab standards are, what equipment is used or how it's tested without buying a a UL publication for thousand's of dollars to find out. The National electrical code (owned by the ANFI) set the GFCI limit,, the UL determines the testing standards on how to determine the limits and the 4th party labs do the testing, and certify the product if it passes. Anyway a little grease will get you a long way in this cascading layer of potential fraud, it's big business.

Looking at the Schurter filter schematics, they are quasi reciprocal. However, with a resistor there to settle resonances, this type of filter should be internal relative to the power switch. In the case of a washer, that switch may be a relay on a board and not amenable to being intercepted on its output. So using such a filter external to the washer would result in bleed current being drawn like a parasitic wall-wart power module.

I'd still try a cap across the line to see if it helps. A 'scope would provide some quantifying support, but blind experimentation might also lead to a favorable result.

Yes the Schurter filter would waste a small amount of power,

This is the response I got from ESP/SurgeX

All of our product filters are bidirectional.

ESP [multistage] shunt suppression works both ways, but SurgeX ASM [Advanced Series Mode] protection works best keeping a surge from going through the input to the product.

Web: ametekesp.com, surgex.com, powervar.com

Since these are designed as external devices - they'd be easier to put in this application.

I expect they have a bit of parasitic load as well.

Thanks. Did you get a filter attenuation chart from them?

Nope. I'll email them today

I've emailed amtek to see if they can provide detailed specs for 2 products.

The ESP Compact Series S1 and the ESP Digital QC D11416T. Both these are 20 amp products that may work. The SurgeX products do have frequency specs, but not listed below 30Khz. The ESP products say contact for detailed info (they say call - but I tried email instead cause I hate sitting on the phone).

The S1 is ~$30. The D11416T is like $100. So - the S1 is a better option if it works.

S1 https://www.ametekesp.com/esp/compact-solutions/compact-solutions-120-208v

D11416T https://www.ametekesp.com/esp/digital-qc/digital-qc-120-208v

BTW - the Envision products include some very need diagnostics including a s/w scope. But they are $200+ and probably not for non-engineering/electrical tech types

Ill let you know if I get a response.

-Bob

So I went to my building department and talked to the building inspector about this issue and was told that a GCFI outlet was not needed at the washer. The only code violation would be if there was a duplex outlet and only 1 plug was to be used do to the fact that there is an electric dryer which we know takes a different plug. He said the same at a sump pump to only have 1 single plug for the pump only.

Generally, local interpretation takes precedence. However, I think in such a case of a metal frame with water about, that I would go belt and suspenders and tie (bond) the washer chassis to a water pipe or other ground that is known to have low conductivity to the grounding point of the breaker box. In the case of a copper water line, there must be jumpers whenever the line is interrupted by meters, filters, or whatever. The breaker box will usually be grounded to the water line input to the house, if municipal.

My 'local' inspector - actually State Electrical Board required it - I had to add a GFCI.

kaseki couple of questions. What gauge wire would be appropriate to use? Now this one might be a stupid question but if a refrigerator has an ice maker (water) and you can never rule out a problem then in your statement above (metal frame with water about) would not or should not a frig need a GCFI outlet or breaker. Same with a dishwasher or are the 2 just different then a washer? Maybe my thoughts are just off track.

First, my SZ 632 (4 ft wide) unit is on a 15A GFCI breaker. I generally use GFCI breakers for most applications that are anywhere near accessible grounds that one might be able to touch while at risk from leakage current. Certain direct wired motor circuits would be exceptions. This is accommodated in the code. The dish washer, garbage disposal, toe-kick heater blower, and anything else powered from below the sink are on GFCI circuits. A review of my breaker arrangement label shows that every circuit in the kitchen is on GFCI except the induction cooktop, induction wok, and wall ovens.

I would have to reread the kitchen part of code to be clear whether I was conforming or being cautious. However in the version of the NEC relevant at the time, which I no longer have, I think GFCI wasn't required above some height such as six feet, and I think the refrigerator receptacle is lower than that.

Wire size is a subject that occupies a large portion of the National Electrical Code, and its counterpart NEC Handbook that anyone thinking of messing with his house wiring should have handy and study. Nominally, for a 15A circuit, 14 AWG conductors within a 2 conductor + ground cable would be used. My kitchen is at least 50 ft from the relevant breaker box, and my reno involved adding conduit fed wireways into the basement to avoid a horrible thicket of Romex on beams and joists. Generally, I wired each circuit that handled a significant portion of its rated current one step higher in wire gauge for the path from breaker box to the interface with the drop from the walls above. So for this circuit, 12 AWG would likely have been used. The object would be lower path resistance, and less required derating for the bundle of conductors in the 1-1/4 conduit portions.

If the question is my SQ washer, the circuit is 20A GFCI, and would nominally use 12 AWG cable. In my case I might have used either 12 AWG or 10 AWG within the wireway. The drop to the washer receptacle is 12 AWG Romex.

So in most cases (somewhat influenced by temperature), the washer circuit should be 20A using 12 AWG cable, but I don't think that there is any prohibition against a 15A circuit for this case, so long as the washer is the only device on the circuit. (I would read that part of the code before doing anything, of course.)

In regards to bonding the washer to water pipe what gauge wire would appropriate? I've seen very heavy gauge wire used to ground the panel and jump water meters and pressure regulators.

The wire size for this purpose has two drivers: Actual likely current before a circuit breaker disconnects, and resistance to damage failure. Hence for the latter driver the 6 AWG solid conductor used in swimming pool area bonding to avoid hazard currents in the microamperes. Breaker boxes need to bond the neutral to various grounding schemes as well as to the utility drop. So their grounding wire sizes depend on various factors.

To the point: While a 14 AWG conductor would be sufficient, I would probably use 12 or 10 AWG, and given potential motion and movement, use green insulated stranded with crimp devices on the ends. One end might be a screw hole lug for attachment by a washer panel screw; the other might be a crimp sleeve that would fit into a grounding clamp that fit the target water pipe. You should use a proper ratcheting crimp tool for most crimp applications.

Note to home owners: If you want to address every possible modification that you might make over the next 20 years, you will likely end up with a lot of tools (+++) and a lot of space taken up with rarely used adapters and fittings (---).

kaseki

"So in most cases (somewhat influenced by temperature), the washer circuit should be 20A using 12 AWG cable, but I don't think that there is any prohibition against a 15A circuit for this case, so long as the washer is the only device on the circuit. (I would read that part of the code before doing anything, of course.)"

The code is silent on the size of the circuit but would require that the wire match the breaker - 20amp needs 12-2w/g. 15 amp needs 14-2w/g. The driving factor on the circuit size would be the appliance draw in this case. The code also does not say "dedicated circuit"- that is from the appliance manufacturer. They recognize that the load of a washer is higher than 'typical outlets" so they don't want to share. Modern washers do have heaters - so they take some power.

Both the GE and the Samsung washers I worked with said

"Connect to a 15 or 20 amp dedicated circuit". So they would both operate on 15 amp circuits. I think most installs will be 20 amp. Mine is.

kaseki

I never got a reply from amtek on the detailed specs for surgex filters.

- Bob

Thanks for the update.

I recently installed this type of breaker for my entire house. It is a Square D homeline Dual function GFCI and AFCI combine breaker. I had 2 issues that I encountered that were both my fault. 1st, I didn't have the neutral(white wire) goes into the breaker as instructed, so when I turn on a high amperage appliance it would trip the circuit. If your main panel is a SquareD QO with plug-on-neutral design, then you probably don't have to worry about wrong wiring. If you have just a QO panel, then check the pigtail and neutral wire for the breaker to make sure they are connected properly and snug. The 2nd issue I ran into was I had an outlet that has the ground wire almost touching the hot side when the outlet is pushed in to the box, wires got jammed together and pushing on each other. So I'd check that wire connection to the receptacle to make sure no wire is connecting. Also, if you use this dual function breaker, you can't use a GFCI outlet in top of it, they won't work together. Hope it helps.

kaseki

Just got this reply from amtek. The apologized for the delay - said the request got lost in the holiday rush.

Since we don't know the source of the noise (if that is what is tripping them), I don't know if these filters would be effective. It would be fun to try - but since my Samsung is working just fine (as opposed to the GE). I would go with the S1....

Both models have a VPR of 330V in all modes, conform to UL Std 1449 and are certified to CSA Std C22.2 No. 8.

D11416T:

S1:

Bob

Thanks. Recorded for future possible use. Too bad they didn't supply graphs.

In case there are any other laypeople out there following this thread, just wanted to provide another update. I finally had the service tech from the appliance come out to look at my (new, replacement) washer that still trips the circuit and he said that nothing is wrong with the machine. I'm wondering if it would make sense for the service tech to actually observe a load of laundry go through so he could see what is happening when the machine trips the circuit.

My next step is to hire an electrician who can hopefully help me resolve this issue so I can wash laundry without having the machine cut off once or twice per cycle.

So Frustrating! Wish I could help more. I was pretty much 100% sure the tech would point the finger. The Electrician may not, but may not have a solution if it is the washer.

We are having this same issue right now in our new build but it is a Maytag. Same thing, electrician says nothing is wrong and we have someone booked to look at the washer on Tuesday! Please help!

A few muses:

This might be a bit nuanced re NFPA 70, but I wonder whether one could be compliant by using a dedicated "motor" circuit (no receptacle outlets) with flex between the box and the washer, and a plain breaker.

I would first buy one of those filters @Robert Gann found and experiment as that would be the approach requiring the least amount of my time, but for an electrically informed DIY-er with time available a dedicated "motor" circuit might be less expensive to incorporate.

In any case, whether adopting filtering, or having to spend time with a series of "investigators," one is using up resources that are entirely due to inadequate filtering in the washer. Maybe replacement with a Speed Queen will solve the problem -- another expensive experiment.

Clever way to get around the code requiring GFIC protected plugs in the laundry area.

Technically - since it is hardwired, there is no plug and you don't need the GFCI

Of course, since you are defeating the purpose of the code - to protect you - you are loosing that protection. To maintain that protection - you'd need to have a GFCI breaker - and if you are doing that, I would just start with a standard plug and a GFCI breaker. If that does not work, probably this would not work.

It might be interesting to see if the hardwired version works better than a GFCI Breaker/Plug approach. I don't know why it would unless the hardwired version provides some filtering the non-hardwired version does not. I can't see why it would electrically. Maybe using shielded cable from the washer to the wall would help(??) - see below.

If you were to do something like this, you need to be careful about wiring and connections.

- The cord(wire) required to run into the washer and the wire in the wall are incompatible. You can't run Romex or armored cable into the washer and you can't run flex through the walls. So you'll have to make some connections. They need to be done correctly.

And by the way - I'm not a licensed Electrician - I'm an EE and not a licensed one, so this is just friendly advice, not authoritative. Check with local code or an Electrician.

If you were to do this;

- Use code appropriate wire from the breaker box to the wall box

- I would use armored cable from the wall box to the washer (the flexible metal wrapped cable)

-> Use the appropriate cover plate and clamp at the wall box location and make sure grounding is correct.

-> When you connect the armored cable to the washer, the connection needs to be legal which means you probably either need to use an integrated 'box' the washer (if it has one) or install a box on the outside and run through a proper hole to connect to the washer. This is very interpretation based... Probably, if required, it has an integrated one.

- Use a GFCI breaker.

BTW - you're probably invalidating the warrenty.

Of course, you can simulate all this with:

-> put standard plug in the wall

-> replace the breaker with a GFCI breaker.

-> If that does not work - this elaborate wire-around probably does not work.

But it is an interesting experiment. If I was desperate, I might try this before putting on the filter discussed. Putting in the filter is going to require adding boxes and stuff anyway.

-bob

Just use a 15 amp GFCI wall plug with a 20 amp AFCI breaker out in the box and you should be fine. That's what did it for me after weeks of jacking around with all other suggestions.

Why 2 GFCI's? The one at the breaker box is doing nothing.

Sight may be being lost of the issue here -- washer motor control generated high frequency noise interfering with the breaker behavior. One can filter the noise or one can use a breaker not affected by it. Filtering is the best way to go, but since (a) the victims of bad washer design and/or bad breaker design are adverse to or unprepared for electrical experimentation, (b) many attempts have already been tried without success using different breakers conforming to the code requirements for laundry room outlets, and (c) the victims may not wish to incur the expense of replacing the washer with a different model, then what is left as a 'last straw' approach is eliminating the GFCI breaker altogether, and that requires careful conformance to code requirements where such a configuration may still be allowed and, in the context of the particular conditions, deemed safe. I haven't investigated the exact wording of the code (it may depend on which year's code is applicable) to see what wiggle room is potentially available.

You are absolutely correct kaseki. I solved my problem by replacing my new GE (both of them) with a Samsung. But I had the benefit if great Lowes customer service that allowed me to return it (them). Still - I spent a couple hundred bucks (per irs mileage deduction) driving to/from various Lowes and at least 12 hrs (and 2 backs) installing/removing in a basement stacked configuration.

I suspect joeb333 managed to find a GFIC (plug) that is not susceptible to the noise from his washer. Perhaps he could share the brand/model/lot# of it.

I would love to see somebody try the filters - I would do it - but since my Samsung is not tripping the GFI - I can't.

My current GFCI wall plug was the same that was in there when the washer was tripping the GE 20 amp GFCI breaker.

I simply ran across an article that stated there were issues with using the 2 GFCI's in the same circuit.

So me not wanting to go with a plain breaker if I didn't have to, installed a GE 20 amp AFCI breaker and the machine has ran like a champ since.

I don't know how or why, I just know there's peace at our casa once again!! :)

Ah- I saw AFCI breaker and read GFCI breaker. I have a conventional (not AFCI) breaker in my system. You are correct, two GFCI's are problematic. Glad there is peace.

The 2014 NFPA 70, National Electric Code, Article 210.12(A)(1) - (6) list all areas that are required to be Arc Fault protected (AFCI), Article 210.11(C)(2) require the laundry receptacle to be fed from a 20 Amp branch circuit. Article 210.8(A)(10) requires all 15 & 20 amp receptacles installed in laundry areas to be GFCI protected. General lighting in the laundry room is permitted to be on another circuit per Article 210.11(C)(2). This is the current code cycle being used in the state of PA. in the end, the (AHJ) Authority having jurisdiction (state or local government) has the ability to omit certain articles of the code. In PA, the AHJ did not omit any code articles. What we have found to fix the issue is exactly what joeb333 stated above. Install a standard AFCI breaker in the load center and a 20 Amp GFCI receptacle at the clothes washer. Thank goodness It's been working like a charm with all makes and models.

HU-885141950 This configuration:

What we have found to fix the issue is exactly what joeb333 stated above. Install a standard AFCI breaker in the load center and a 20 Amp GFCI receptacle at the clothes washer. Thank goodness It's been working like a charm with all makes and models.

Does not work in all cases. Sometimes the opposite works - GFCI breaker and not GFCI receptacle. Both configs have worked, and have failed. Seems to be install specific - which is not surprising if the hypothesis of VDM noise injection is correct.

Glad it is working for you. For me the solution was replace the GE washer with a Samsung.

Luckily I was able to do that because it was 2 (two) brand new GE washers that I could return under 30 day policy. The Samsung never looked back.

Glad to hear that there seems to be a fix to this issue. I need a Square D AFCI but all I can get my hands on is a CAFCI which is supposed to be a newer version of the AFCI with extra protection. So the question is will this extra protection (versus AFCI) continue to have the same issue that has been happening or simply more protection?