40 Volts on dishwasher frame, circuit board design issue?

[tortuga]

Brand new dishwasher on a kitchen remodel trips the GFCI, its a 12/3 MWBC, my coworker (electrician) swapped gfci's and used a extension cord to go to another GFCI on another circuit.
Trips all of them.
Dishwasher came with factory wired cord.
Next appliance warranty service comes out says faulty wiring, not sure what he did, homeowner suspected nothing.
Contractor requested a GFCI breaker but I forgot it was a MWBC, that would be an expensive breaker.

Today I visited the site for the second opinion and I plugged in one of those 2-3 prong adapters and dishwasher ran fine.

I measured 40v from the grounding tab on the plug to the metal outlet box, with a good multi meter true RMS.
I used my good old wiggy for the same test and it did not light up or vibrate, so it must be a really low current.
Ran a known good ground wire to the panel to test still 40 Volts on dishwasher frame.
I did not try to ammeter as I dont think my meter would be accurate in miliamp range.
I realize I need one of those adapters that lets you read the miliamps on the cord, and go meet the appliance guy there.

To you electronics / engineers on here I realize there are surge protectors you put on your circuit boards with some kinda blue caps connected to our EGC, it there any situation it would be 'normal' for there to be 40 volts on an open ECG like that?
Thanks in advance

 

[hbiss]

I realize there are surge protectors you put on your circuit boards with some kinda blue caps connected to our EGC,

I think you nailed it. Your meter is high impedance, hence the 40V reading. Put a little load on it and it goes away. But the GFCI doesn't put a load on it either and there is enough current flow to equal at least 5mA. What to do about it? Do the instructions say anything about a GFCI protected circuit? I'm wondering if it was designed not to be used with GFCIs. Either that or it's defective.

-Hal

 

[synchro]

I did not try to ammeter as I don't think my meter would be accurate in miliamp range.

Are you referring to a clamp meter or an inline ammeter?
You could get ten times more sensitivity from a clamp meter by wrapping 10 turns of the ground wire around the clamp. Whether that's enough to get meaningful readings would depend on your meter.

 

[LarryFine]

Today I visited the site for the second opinion and I plugged in one of those 2-3 prong adapters and dishwasher ran fine.

I measured 40v from the grounding tab on the plug to the metal outlet box, with a good multi meter true RMS.
I used my good old wiggy for the same test and it did not light up or vibrate, so it must be a really low current.

I would next have the unit running, and jump the ground tab to the outlet box. If it trips, it's the dishwasher.

Does it trip when plugged in, or only when turned on? If the former, I'd open the junction box and take a look.

 

[don_resqcapt19]

I think you nailed it. Your meter is high impedance, hence the 40V reading. Put a little load on it and it goes away. But the GFCI doesn't put a load on it either and there is enough current flow to equal at least 5mA. What to do about it? Do the instructions say anything about a GFCI protected circuit? I'm wondering if it was designed not to be used with GFCIs. Either that or it's defective.

-Hal

The NEC has required the outlet that supplies a dwelling unit dishwasher to be GFCI protected since the 2014 code, so it should be designed to work with GFCI protection.

 

[kwired]

Are there any dishwashers with electronic motor controllers out there? I'm not familiar with any but wouldn't be surprised. I have had problems with clothes washers with such motors/controllers not playing well with GFCI protection.

One I did resolve a problem with has a ground isolation switch that closes the EGC to the motor frame within the appliance if the back cover of the appliance is removed. Plastic mounting clip on the switch assembly broke and switch "dismounted" allowing the contact to close. Any time the motor attempted to start it would trip the GFCI.

One of my first things I did while troubleshooting this one was to put a two to three wire adapter on the appliance cord like in the OP here- it would hold while on the adapter - so I knew it had to be something faulting to the EGC within the appliance.

 

[don_resqcapt19]

Are there any dishwashers with electronic motor controllers out there? I'm not familiar with any but wouldn't be surprised. I have had problems with clothes washers with such motors/controllers not playing well with GFCI protection.

One I did resolve a problem with has a ground isolation switch that closes the EGC to the motor frame within the appliance if the back cover of the appliance is removed. Plastic mounting clip on the switch assembly broke and switch "dismounted" allowing the contact to close. Any time the motor attempted to start it would trip the GFCI.

One of my first things I did while troubleshooting this one was to put a two to three wire adapter on the appliance cord like in the OP here- it would hold while on the adapter - so I knew it had to be something faulting to the EGC within the appliance.

Even if there are, there is no excuse on the part of the manufacturers, as unlike other equipment that have had very recent code requirements for GFCI protection, dwelling unit dishwashers have had that requirement since the 2014 code and the manufacturers are aware of that.

 

[LarryFine]

You liked this question, but didn't answer it:

Does it trip when plugged in, or only when turned on?

 

[tortuga]

Thanks everyone for the replies

You liked this question, but didn't answer it:

Unfortunately I did not think to test that while I was there, I will when i go back.

Are you referring to a clamp meter or an inline ammeter?

This is my current meter

1000A AC/DC TRMS TightSight® Clamp Meter

61-775: TightSight second display Tapered jaws with hook tip True RMS current and voltage AC/DC voltage Audible continuity High frequency rejection* Peak hold Max/min hold Data hold Selectable audible and visual alerts when potentially hazardous voltage

www.idealind.com

perhaps I should get one of these for when i get called out to these kinda things:

https://www.tequipment.net/HT-Instruments/LineSplitter/Line-Splitters/?Source=googleshopping&gclid=CjwKCAiA9tyQBhAIEiwA6tdCrKf6KLA-1kekTw2kfE1o0PdWh2A_8WFfqwz6Cu3a8kXAPoI4LTSUgBoCPdYQAvD_BwE

 

[kwired]

Even if there are, there is no excuse on the part of the manufacturers, as unlike other equipment that have had very recent code requirements for GFCI protection, dwelling unit dishwashers have had that requirement since the 2014 code and the manufacturers are aware of that.

I agree. My case I described they apparently did find a way to deal with the issue, that method just happened to fail. You don't find much information about it on web searches so far either. The washer I was dealing with leaves the motor/transmission components "unbonded" as long as the back cover is installed on the appliance (and switch mounting method doesn't break). Bet many appliance repair guys don't understand this and will tell owner there is a wiring problem or even that the machine is not compatible with GFCI's, when in reality what I found was a simple malfunction and easily repairable.

Even before these kinds of appliances that may not play so well with GFCI's I've had freezers in garages or unfinished basements that trip GFCI's and appliance guy tells owner the GFCI is defective. I come and find there is a ground fault and GFCI is doing what it should do. Then have to look into the appliance to prove I was right and appliance guy is wrong. So far am 100% at finding "predictable" trips. The trips that only occur say couple times a year are probably from outside surges or other interference that is not so regularly occurring.

 

[hbiss]

The NEC has required the outlet that supplies a dwelling unit dishwasher to be GFCI protected since the 2014 code, so it should be designed to work with GFCI protection.

Agreed. But did the manufacturer get the memo?

This is my current meter

perhaps I should get one of these for when i get called out to these kinda things:

That adapter will only allow you to measure the current on the hot and neutral and there may well be things in the machine that draw current even when the machine is off. You want to measure the current on the ground while connected to a non-GFCI protected receptacle. What you can do is break out the green ground with something like a receptacle/plug connected together with stranded THHN. Leave the green extra long and wrap it around your meter jaws like 10 times. Then divide the reading by 10. That's the way that adapter works by the way so save your money.

-Hal

 

[synchro]

This is my current meter

1000A AC/DC TRMS TightSight® Clamp Meter

61-775: TightSight second display Tapered jaws with hook tip True RMS current and voltage AC/DC voltage Audible continuity High frequency rejection* Peak hold Max/min hold Data hold Selectable audible and visual alerts when potentially hazardous voltage

www.idealind.com

perhaps I should get one of these for when i get called out to these kinda things:

https://www.tequipment.net/HT-Instruments/LineSplitter/Line-Splitters/?Source=googleshopping&gclid=CjwKCAiA9tyQBhAIEiwA6tdCrKf6KLA-1kekTw2kfE1o0PdWh2A_8WFfqwz6Cu3a8kXAPoI4LTSUgBoCPdYQAvD_BwE

View attachment 2559598

That allows you to measure the differential leakage current on the L and N pair (like a GFCI does), or directly on the ground conductor. It offers a 10X multiplier but only for the line current on the hot conductor, not for the leakage current.

Your clamp meter has at best 100 mA resolution, so I don't think it's going to be adequate for measuring leakage currents that may be causing GFCI tripping --even if you have a 10X multiplier coil.
Someone recently posted a clamp meter that could measure relatively low currents with a reasonable price, but it wasn't a familiar brand name. I'm thinking it may have been Hv&Lv, but I may be wrong about that.

 

[kwired]

Agreed. But did the manufacturer get the memo?



That adapter will only allow you to measure the current on the hot and neutral and there may well be things in the machine that draw current even when the machine is off. You want to measure the current on the ground while connected to a non-GFCI protected receptacle. What you can do is break out the green ground with something like a receptacle/plug connected together with stranded THHN. Leave the green extra long and wrap it around your meter jaws like 10 times. Then divide the reading by 10. That's the way that adapter works by the way so save your money.

-Hal

I'm not so certain. upper right marked mA PE? does that possibly mean you can measure the PE (EGC) with that section? Never seen such a thing before but can see it being possiblilty, they maybe ran both other conductors through other side of that upper section and if so any reading there would be any imbalance between the two conductors.

 

[tortuga]

Just checked out the UL standard for dishwashers, and for leakage current it just points to UL 101.
UL 101 has some strange tolerances for hi frequency leakage current, they call it a measurement indication unit or MIU.

 

[hbiss]

UL 101 has some strange tolerances for hi frequency leakage current,

Probably when a VFD is used. Lots of instances of them tripping GFCIs. A/Cs and Pentair pool pumps come to mind.

-Hal

 

[Mr. Serious]

a clamp meter that could measure relatively low currents

Alternatively, you could use a regular old non-clamp multimeter, that has an AC mA range. Connect the probe in the mA position and run the ground current through the meter. If you have one of those somewhere, you wouldn't have to buy a new meter for this.

 

[ptonsparky]

Alternatively, you could use a regular old non-clamp multimeter, that has an AC mA range. Connect the probe in the mA position and run the ground current through the meter. If you have one of those somewhere, you wouldn't have to buy a new meter for this.

Safer for the meter would be to series a low resistance in the ground, measure the VD and use a bit of ohms law.

I usually do it your way and risk the meter fuse after I'm relatively sure the current is within the meters range. Only replaced the fuse a few times. Keep a spare on hand.

 

[hbiss]

I usually do it your way and risk the meter fuse after I'm relatively sure the current is within the meters range. Only replaced the fuse a few times.

Rule is, if you are measuring an unknown current or voltage, start with the highest range and work your way down until you get to the one that indicates the correct value.

-Hal

 

[tortuga]

Alternatively, you could use a regular old non-clamp multimeter, that has an AC mA range. Connect the probe in the mA position and run the ground current through the meter. If you have one of those somewhere, you wouldn't have to buy a new meter for this.

Interesting I have a few meters with that series mA position on my home 'electronics bench', one is a old simpson 260 the other is a newer one, I have never used that function will play around with it.

Safer for the meter would be to series a low resistance in the ground, measure the VD and use a bit of ohms law.

Thats a interesting idea, I do have some chunky ceramic 1 ohm 5 watt resistors

I never knew about UL 101, its aimed to be a almost universal standard for appliance leakage current, and it has a pretty complete schematic of how they build their 'leakage current meter' with just a few components and a regular mA meter.

I have not been back to do more tests yet, will post an update when I do.

 

[ELA]

I plugged in one of those 2-3 prong adapters and dishwasher ran fine.
View attachment 2559590


To you electronics / engineers on here I realize there are surge protectors you put on your circuit boards with some kinda blue caps connected to our EGC, it there any situation it would be 'normal' for there to be 40 volts on an open ECG like that?
Thanks in advance

The condition you created by lifting the EGC is the abnormal condition that produced the 40V. If the manufacturer uses capacitors for EMI filtering that can then create a voltage divider from Neu to Line w/respect to Gnd. When Gnd is lifted. If the Neu to gnd and line to gnd caps are exactly the same value you would read 60V.
The small value caps used are relatively high impedance so current flow is lower than the 5ma trip point under normal conditions.