Thread: GFCI issue corrected, but has complaint

A recently installed new listed GFCI in bath ,and also, had installed an odor fan {Doane}
also listed.
the GFCI would trip using the trip/reset button, but would not trip using a tester[ either an
Ideal or Etcon] : With some testing , the problem was isolated to a one and a half ohm
resistance on the ground wire in the main panel. The ground wire was cut short,and had
a twisted jumper to the buss bar. By using a proper manual connector , the problem was
cured, and the testers did trip as required.
Now , intermittently ,the GFCI will trip when shutting OFF the fan only, but only real
slow or real fast shut off, normal easy speed seems OK. The switch is a hospital
grade -off / on -switch , and is wired to the load side of the GFCI.
There seems to be some isolated " loose" current involved. Any suggestions ?
Or , should the fan be wired to the line side of the GFCI ?

" when experience and wisdom have been gained , youth has been paid. "
by Jack London

Similar Threads:

• rafter issue
• Back to the collar tie issue...
• Sewage ejection pump-GFCI?
• Testing GFCI Outlets
• Kitchen area GFCI

Originally Posted by Stephen DeCosta

The switch is a hospital grade -off / on -switch , and is wired to the load side of the GFCI.
There seems to be some isolated " loose" current involved. Any suggestions ?
Or , should the fan be wired to the line side of the GFCI ?

The exhaust fan should not be wired from the bathroom receptacle outlet circuit.

An exception would be if that receptacle outlet circuit ONLY supplied that one bathroom.

Wiring off the load side of the GFCI would indicate a probable problem with the fan.

Wiring off the line side of the GFCI would hide a probable problem with the fan.

But it really should not be wired from the GFCI circuit anyway - not if installing to code.

Regular 'residential' grade snap switches have a 'mushy' snap and are often rated to be used as motor (fan) controls to 80 percent of their amp rating for other (usually 'T' rated - T being tungsten as in lamp filaments).

'commerical', 'industrial' snap switches have a clean 'non-mushy' SNAP and are often rated at a MUCH LOWER motor (fan) control rating - usually around 50 percent sometimes less. They are meant to be cleanly and directly (often loudly) snapped on or off.

Hospital grade are usually clean (loud) 'non-mushy' snaps with isolated ground.

Do not rely or assume "self grounding" even if using a metalic/ferrous metal box. Use a grounding wire, and isolate the switch from the metalic box using the square washers.

Arcing in the switch if it is operated too slowly and not fully opened or closed.

I'd be looking for a bootleg ground, nail pierce, damaged wiring insulation, etc. Most likely DIY error is to miswire a switch loop. If power from the panel was at the fan - etc.

You didn't say what the wiring was, i.e romex, AC, emt, etc.

Also if the circuit is a multi-wire branch circuit, if the circuit is serving anything other than the bathroom, etc. could be of issue.

If you don't know what you're doing, or what you did; and are unfamiliar with circuit tracing, diagnostics, etc. an electrician would be your best bet.

Finally, you would need to confirm the date of manufacture of the combination GFCI receptacle. The standards have changed several times, there was and is no prohibition for the older types to be sold. You would want to assure the one you installed meets the newest standard requirements. From the symptoms you described, it does not seem you COULD have installed one that meets the newest standard requirements.

Steve

First good job finding the issue with the equipment ground wire not spliced correctly in the panel. The tester needs an equipment ground to bleed current to in order to fool the gfci.

The gfci however does not need the equipment ground to function properly ... but you certainly improved the safety of the branch circuit by finding that faulty splice. You likely had no protection from an overcurrent event like a hot wire to ground .. fault.

Intermitent tells me that you do not have a condition like a permanent connection of a neutral to ground or the like. If you did the gfci would trip as so as you turn the fan on. Or if it was a multiwire circuit wired incorrectly with a gfci receptacle. It would trip immediately on turning fan on.

However the gfci should not be affected in any way shape or form protecting that exhaust fan. So I'm thinking you might have an issue in the wiring compartment between a ground wire and a neutral in the fan wiring compartment. I'm not exactly sure how yours connects to the branch circuit..some actually have a cord and plug and you wire the receptacle and some you wire to insulated motor leads. Any way i would take a look and see if you have any close clearances between a neutral and ground at the fan or even the switch box. Could be vibration or just some oddity that is causing neutral to contact ground. Could be a bad gfci but that is not what I would suspect...it is probably doing its job. Lastly some motors just have too much leakage of current or are borderline as to what is required and will trip a gfci once and a while. I don't really know the physics of it but it usually trips the gfci turning on or off the motor.

The switch is another possibility....easy to try a new switch. Generally speaking if the house is not brand new it would not be uncommon to have gfci receptacles and fans lights on the same branch circuit in a bathroom. The circuit should be a 20 amp one that has been a requirement for many years as has been gfci for bathroom receptcles. I wouldn't say it is uncommomn to have the exhaust fan gfci but they usually are not gfci protected nor are the lights. I'm not saying wire to the line side of the gfci...especially now that you say you have a gfci tripping out. I'd try to find out why it is tripping.

I do agree with Jerry that if possible I'd get the fan off the receptacle circuit. Not the end of the world but it will provide more power to use for those mega turbo hair dryers so the breaker doesnt trip... .. sounds like your not having those kinds of issues though.

cheapest/easiest first is what I'd do..... replace the switch, then the GFI, then the fan (or at least try another fixture so you can rule out the current one as the problem).

After that it gets a bit more time consuming but not really too difficult. Just work logically through it and you'll find the problem.

Originally Posted by Roger Frazee

Generally speaking if the house is not brand new it would not be uncommon to have gfci receptacles and fans lights on the same branch circuit in a bathroom.

I do agree with Jerry that if possible I'd get the fan off the receptacle circuit.

Roger,

The reason I pointed out that the exhaust fan is not allowed on the receptacle outlet circuit in the bathroom (with exceptions) as because ... ...

Originally Posted by Stephen DeCosta

A recently installed new listed GFCI in bath ,and also, had installed an odor fan {Doane} also listed.

He "recently installed" the GFCI (I took that as meaning he recently installed the receptacle) "and also, had installed" the bathroom exhaust fan, which "recently installed" means they should not be on the same circuit (with exceptions).

Yes, the exhaust fan motor *could* be causing the problem, but like you, I doubt it.

I have heard of switches causing GFCI to trip, due (I have heard) to the arcing during switching. That should not affect a GFCI, so I suspect there is another intermittent cause which is unknown, that it is the user reporting that the problem is "the switch" and not what is really the problem causing the GFCI to trip.

By the way, welcome to the board.

Thank you for all the help. I will give more detail on the issue.
The wire is NM -- 12/2 with ground, connected to a single outlet
in living room and single outlet in spare room, Ground wires measure
"0" OHMs from fan motor through switches, GFCI, Boxes,and back to
panel. With power off and disconnected at the panel, ground to hot measures
infinite , ground to neutral measures infinite, and neutral to hot measures
infinite. The GFCI is LEVITON, [CAT 78999-1] 20 amp rating -- the
switch is G.E. co ,U.L. INC, 20 , FED. SPEC. - W-S-896d , with
20 amp rating.
The trip never happens when turning on fan. nor any other situation except
occasionally when shutting fan off. The Fan itself appears to be working
properly , and it was tested for continuity prior to installation , mainly, just
as a practice .
As to the discussion , It may very well be uneven current in the switch
itself. more work .

I would be very surprised if it is your switch, I've never quite been able to get my head around why a switch would trip a gfci. I have heard of this but it my experience I've never had a switch be the root problem to a intermittent gfci trip.


I few years ago I bought some Leviton gfci's in a 10 pack, everyone of them was so temperamental it would trip if you looked at it wrong. I could take my strippers and tap the cover plate and the dang thing would trip just like if you pushed the test button. Just saying there is such an animal where the trip setting of a gfci might be a tad to sensitive.


Jerry

I actually knew why you said that about the fan on the receptacle circuit but for some reason came up with a reason of my own ... go figure..

Thanks for the welcome

I got off to a pretty shakey start here so I offer my apologies for some very poor judgement on my part.

Originally Posted by Roger Frazee

I would be very surprised if it is your switch, I've never quite been able to get my head around why a switch would trip a gfci. I have heard of this but it my experience I've never had a switch be the root problem to a intermittent gfci trip.

Same here, which is why I said "(I have heard)" ... I have heard that ... but darned if I have been able to figure out the how it would trip a GFCI.

I few years ago I bought some Leviton gfci's in a 10 pack, everyone of them was so temperamental it would trip if you looked at it wrong. I could take my strippers and tap the cover plate and the dang thing would trip just like if you pushed the test button. Just saying there is such an animal where the trip setting of a gfci might be a tad to sensitive.

May have been the same ones which would reset with with reverse polarity?

I showed a Leviton representative about that problem, could almost do it repeatedly ever time by pushing the reset button in on the left side of the reset button instead of the center of reset button - they never could figure the how it would reset reverse polarity - but it would. That was before the change to the lock out style which prevents line/load reversal. Have not had any reset reverse polarity with the new ones.

Originally Posted by Jerry Peck

Same here, which is why I said "(I have heard)" ... I have heard that ... but darned if I have been able to figure out the how it would trip a GFCI.



May have been the same ones which would reset with with reverse polarity?

I showed a Leviton representative about that problem, could almost do it repeatedly ever time by pushing the reset button in on the left side of the reset button instead of the center of reset button - they never could figure the how it would reset reverse polarity - but it would. That was before the change to the lock out style which prevents line/load reversal. Have not had any reset reverse polarity with the new ones.

Ya know I heard that about the levitons and always wondered if maybe that is what I had got my hands on. Basically I ended up just pitching them.

An interesting thing about gfci's you may or may not know is that an overcurrent fault may or is likely to not trip a receptacale gfci. I watched this done at an electrical trades gathering by a tech from Pass and Seymour Legrand. He used a light switch to impose a hot to ground fault on a receptacle gfci. The circuit breaker tripped but the gfci never did...he repeated this several times. What did happen though is you could see arcing in the gfci everytime he placed the fault into the circuit. The gfci would not test or reset after the second ground fault. So they do not like overcurrent faults at all. And likely will not function properly after having that kind of amperage passed through them. I'm not saying they won't but there is a big chance they will not trip in this kind of situation.

Another thing that some people don't believe is that RF will trip a gfci. In a warehouse one time the maintenance guys showed me how they could hold their hand radios within four feet of the gfci outlets at the PLC cabinets and key them and that dang gfic would actually start vibrating and then trip out. Wouldn't have believed it till I witnessed it for myself.

Originally Posted by Roger Frazee

An interesting thing about gfci's you may or may not know is that an overcurrent fault may or is likely to not trip a receptacale gfci. I watched this done at an electrical trades gathering by a tech from Pass and Seymour Legrand.

Not surprised at all.

That is because a GFCI *is not supposed to* trip on an overcurrent condition.

The GFCI trips on an unbalanced current between the hot and the neutral, without regard to how high that current is. As long as the same current is in the neutral and the hot, the GFCI does not detect any unbalanced current.

He used a light switch to impose a hot to ground fault on a receptacle gfci. The circuit breaker tripped but the gfci never did...he repeated this several times.

That part throws me, though, as that creates the unbalanced current the GFCI sensing coil is monitoring. Not sure why that did not trip the GFCI switching action ... unless it had something to do with the sudden and real high surge in current which somehow locked the GFCI device up before it could trip off?? That part as me stumped.

What did happen though is you could see arcing in the gfci everytime he placed the fault into the circuit. The gfci would not test or reset after the second ground fault. So they do not like overcurrent faults at all. And likely will not function properly after having that kind of amperage passed through them. I'm not saying they won't but there is a big chance they will not trip in this kind of situation.

Almost confirms my suspicion that the GFCI tripping mechanism somehow 'locks up' with that sudden high current surge and does not trip. Whatever caused it to 'lock up' (maybe something deformed internally?) would logically not allow it to work properly afterward.

Another thing that some people don't believe is that RF will trip a gfci. In a warehouse one time the maintenance guys showed me how they could hold their hand radios within four feet of the gfci outlets at the PLC cabinets and key them and that dang gfic would actually start vibrating and then trip out. Wouldn't have believed it till I witnessed it for myself.

Easy to understand.

The GFCI works by having the neutral and hot conductors go through a coil (very basic description of them here) and the offsetting magnetic fields of the same current flowing in opposite directions cancel out, leaving no magnetic field. The unbalanced current from a ground fault creates a difference between the two currents, which creates a difference between the two magnetic fields, which leaves one magnetic field stronger than the other, which creates a current flow in the coil which, when strong enough, trips the GFCI internal switch mechanism (calibrated to trip at 4 ma to 6 ma ... and which should trip anywhere above 6 ma, but apparently 'locks up' under high current and the resulting high magnetic field created by that high current).

The RF signal is nothing more than electromagnetic waves being broad_cast (had to insert the underscore to defeat the auto text link) through the air, they interact with the coil, and with the right core, the right number of turns around the coil, and the right wave length of the RF and ... a current is generated in the coil and the GFCI trips.

Many, many, many years ago I worked with microwaves, which are also only electromagnetic waves, except that we kept them controlled inside wave guides until they reached the point at which we measured them, and in our case we pumped them into heat sinks which were either air or water cooled and measured the temperature rise to calculate the wattage (mW, W, KW, or MW) power output. In real life they would output into the air through the antenna, the larger ones being used for radar on battleships and ground based installations.

Yes, I can completely understand how the RF radios would cause a GFCI to trip.

But I do not understand (I have a guess, though) why the hot-to-ground fault did not trip the GFCI (because the high magnetic fields deformed the mechanism and caused it to 'lock up').

okay, a little drift here but still somewhat on topic..... ever wonder why your inductive tester doesn't work when you stick it in an empty light bulb socket? The magnetic fields from hot and neutral canel out to zero so there's nothing for it to pick up. Use a jumper wire from the hot and use your tester a few inches outside of the neutral female threads of the light bulb holder and you'll get your reading.

He used a light switch to impose a hot to ground fault on a receptacle gfci. The circuit breaker tripped but the gfci never did...he repeated this several times.

That part throws me, though, as that creates the unbalanced current the GFCI sensing coil is monitoring. Not sure why that did not trip the GFCI switching action ... unless it had something to do with the sudden and real high surge in current which somehow locked the GFCI device up before it could trip off?? That part as me stumped.

It should stump you... because I had a slight brain malfunction... it was a hot to neutral fault not hot to ground..... And yes I agree with your analysis for the reason it didn't trip.

I'm going to have to start getting some sleep after spending 24 hrs getting dug out from this blizzard .... you make too many careless oversites when your brain is in bed and the rest of you is on the puter.....

I figured the RF had something to do with magnetic field in the toroid coils of the gfci but your the first person that gave me an analysis that made good sense.

The circuit was very simple but it was many years ago ... I believe it was like this drawing .. the light showing when the breaker tripped and the gfci we could watch . the energy released actually made the romex he used to construct the circuit move a bit.

Give this a look-see Ground Fault Current Interrupters It's an old article but explains a lot about GFCIs

Originally Posted by Stephen DeCosta

Thank you for all the help. I will give more detail on the issue.
The wire is NM -- 12/2 with ground, connected to a single outlet
in living room and single outlet in spare room, Ground wires measure
"0" OHMs from fan motor through switches, GFCI, Boxes,and back to
panel. With power off and disconnected at the panel, ground to hot measures
infinite , ground to neutral measures infinite, and neutral to hot measures
infinite. The GFCI is LEVITON, [CAT 78999-1] 20 amp rating -- the
switch is G.E. co ,U.L. INC, 20 , FED. SPEC. - W-S-896d , with
20 amp rating.
The trip never happens when turning on fan. nor any other situation except
occasionally when shutting fan off. The Fan itself appears to be working
properly , and it was tested for continuity prior to installation , mainly, just
as a practice .
As to the discussion , It may very well be uneven current in the switch
itself. more work .

There is good reason why bathroom receptacles are supposed to be on dedicated circuits, either serving bathroom receptacles only (multiple bathrooms) or dedicated to supplying multiple outlets for a single bathroom only.

From your update we now know this is not the case with this circuit, which you have now added a fan to.

Older electronics and surge protection power strips often use MOV for transient voltage suppression - thus sending transient over voltage to ground.

An MOV could explain the symptoms you describe (using switch, opening switch loop while motor in operation, causing GFCI to detect an inbalance and to trip).

There are other possibilities by themselves or in conjunction with others.

MOVs (Metal Oxide Varistors) are commonly found in power strips (and other devices) which can provide transient surge protection. MOV circuit protection can send (dump) excess current to ground by circuit design.

An MOV containing device somewhere in the branch circuit (such as a MOV surge protecting circuit in a power strip, a power supply for (usually older) electronic equipment, some fluorescent lamps/fixtures, etc.), and the operation of a snap switch to open the work of the fan motor, could explain the symptoms you describe. There are other possibilities (problem conditions that may exsist) as well.

Keep in mind that every trip condition (outside of intentional test using the integral test button of the device itself) is spending the useful life effectiveness of your GFCI device.


Previously mentioned MWBCs. That particular wiring design called the multiwire branch circuit fools a single GFCI into thinking there has been leakage. An electrician can determine whether any particular circuit is part of a multiwire circuit, at least a legally-installed one. Sometimes ignorant work can create an illegal multiwire circuit (wiring error), which can be both dangerous and hard to identify or correct.

Your addition of this fan to the bathroom and the introduction of the needed GFCI protection for the bathroom receptacle should be accomplished by running the required dedicated 20 amp bathroom circuit. Modification of the original circuit should invoke the latest adopted Code's requirements. Safety is of issue, since this seems to be your own property you have the opportunity to make the election to "upgrade" safety even if your local authority doesn't require it - you should get a permit to do so and have inspected, this would not be considered "maintenance" (new dedicated circuit, or extending - introducing the fan -an older one).

From your original description, and the update, I am still doubful that the date code for manufacture date of the GFCI receptacle is one that would have been produced to meet the latest requirements (changes) to the UL Standard.

P.S. if either of the outlets (living room, spare room) are switched there may likewise be a wiring error as well (ex. switched neutral, etc.). This single "outlet" in the living room and single "outlet" in the spare room - are these lighting "outlets" and if so, switched, and if switched how (snap switches, dimmer switches); OR are they "receptacles"? and if receptacles, simplex or duplex receptacles, switched or unswitched (& what type of switch) ?

You could not begin to diagnose without first removing all cord and plug devices from the entire circuit. I would be suspicious that as described these would be the only "outlets" (locations where electricity can be 'worked'), be them receptacles, lighting, etc., on this circuit, that there isn't a compromise to the wiring anywhere on this circuit (nail pierce, loose connection, floating or bootleg ground, etc.) or a wiring error, if there is no MOV protection circuit somewhere used.

If your "fan" is at a minimum listed "for damp locations" it should be able to be installed on a GFCI protected circuit - how was it labeled and what do the instructions say on those two subjects?

Final question, is this a recirculation only fan not an exhaust fan? is there a possiblity that moisture is building up in either the switch or the GFCI despite the fan being in use?

Originally Posted by Roger Frazee

it was a hot to neutral fault

Now I understand why it did not trip the GFCI.

'Cept that means it was a "hot to neutral" short.

It's a bit chilly here too ... 53 outside this morning ... (but I remember those cold arse winter days of being snowed in as my hometown is about 40 miles south of Buffalo, NY and we didn't move to Florida until I was 10).

'Cept that means it was a "hot to neutral" short.

Ok you got me again...


Bill

Thanks for the link. I'll bookmark it and read through in a day or two. Right now I'm thinking about catching a flight to Florida .....

More info;--Thank you Jerry and Roger . I do understand RF, as I was an ET in
the US Navy many years ago. I do not believe RF to be a problem at this time ,
unless I am being secretly bombarded with RF. There is currently much info
on RF weaponry, as in the site "MK Delta" . Much more sinister is the site
"Project Bluebeam by Serge Monast"
The Fan was purchased at Lowes, and sold as a bathroom exhaust fan. The
box and info had been discarded. Although I can see moisture as being
problematic. How much does it take to interfere with the operation of a GFCI,
I do not know.
The GFCI's were purchased locally in a batch of 8, and I had replaced the suspect
unit with one of the other 8, with no difference in the outcome.
All the outlets in the circuit were tested wire by wire back to the main panel and measured "0" OHMs. and infinite between them.
I did not know what a "MOV" was and looked it up . I do not believe such a device exists on any of the circuits in this electrical system.

Hi Steve

I suspect the fan you purchased was a ' Broan' . Broan fans if they are strictly ventilation fans are acceptable with gfci protection. Nuisance trips are something you fight sometimes to find the cause. If the fan is not installed over the shower or tub and is installed in the ceiling I wouldn't lose a lot of sleep wiring it line side on the gfci. There really is no reason to have it gfci protected unless it is over the shower or tub. Or if your worried I'd try to exchange it for a new one. Seems it is only tripping by action with the wall switch so I just don't see a safety hazard. The primary concern for gfci protection in bathrooms is for the receptacles that you have appliances plugged into that might get wet or dropped in the tub. Or your wet hands etc in contact with the cords and so forth.

I can attest to a switch causing your problem. I have some equipment on a power strip and if I don't *snap* the switch quickly----the GFI trips. I have been meaning to replace that strip. Anyway, I would definitely follow the recommendations to replace the switch first.

Originally Posted by Stephen DeCosta

More info;--Thank you Jerry and Roger . I do understand RF, as I was an ET in
the US Navy many years ago. I do not believe RF to be a problem at this time ,
unless I am being secretly bombarded with RF. There is currently much info
on RF weaponry, as in the site "MK Delta" . Much more sinister is the site
"Project Bluebeam by Serge Monast"
The Fan was purchased at Lowes, and sold as a bathroom exhaust fan. The
box and info had been discarded. Although I can see moisture as being
problematic. How much does it take to interfere with the operation of a GFCI,
I do not know.
The GFCI's were purchased locally in a batch of 8, and I had replaced the suspect
unit with one of the other 8, with no difference in the outcome.
All the outlets in the circuit were tested wire by wire back to the main panel and measured "0" OHMs. and infinite between them.
I did not know what a "MOV" was and looked it up . I do not believe such a device exists on any of the circuits in this electrical system.

I defined an MOV as a Metal Oxide Varistor in that post and described the type of circuit containing an MOV that would be of issue. Electronic devices (mostly older) often contain a surge protection MOV circuit - older TVs, Power supplies, and those notorious "Power Strips". I even explained how those circuits work (dumping transient surges to ground).

These are not "hard wired" to a circuit - simply having a device which contains a MOV surge protection circuit plugged in anywhere on the branch circuit can cause the symptoms you described when opening a switch that cuts off a 1 Ph 120 VAC supplied motor while running (cutting off the power to the motor) older fluorescent, and other such devices. This switch can be the one on the power strip if the power strip is supplying the motor device, or a wall switch (switch loop) supplying as in this case - your fan elsewhere on the circuit. It is common for switch operation supplying a motor to cause a GFCI receptacle to trip - even if the MOV containing device is plugged in elsewhere on the branch circuit on a portion which is NOT being GFI protected. This GFCI trip "MOV" symptom shows up ONLY as a AC Motor (or other time sine shift load) is being SWITCHED OFF that's why I mentioned it.

I have never heard of a Doane branded anything. Nutone or Broan perhaps. Previously described as an odor fan (generally recirculating either with a filter - often charcoal based, or a scent/odor masker) now described as an exhaust fan - unknown if this is even being properly exhausted to the outdoors.

It doesn't take much of any moisture (humidity relative or condensation) present in the wiring seeping in or saturating beneath the cable sheath, a nick in the individual hot or grounded conductor insulation within the sheath (often caused by a stress - too tight staple and a pull, a nail pierce, etc.), at the contacts, collecting on the yoke, or in a fan box, junction box, etc. to cause a ground fault.

If you have no devices plugged in or any dimmer switches, etc. anywhere on the circuit (or have you actually said this?), seems you may indeed actually have a fault situation, or might have a batch which was under recall. Moisture rolling back down in the fan housing even - who knows. Regular NM or NM-B shouldn't be beyond the thermal envelope of the conditioned space - in outside wall cavities or up in a cold attic (esp. with unknown exhaust provisions for new bathroom fan....) equals bad news. NMC is better suited for DAMP locations. There are other preferred wiring methods as well.

You can call the manufacturer of your GFCIs (they have a toll free number) and have them guide you as to where to locate the DATE CODE for date of manufacture of your GFCI devices, This will determine if they are made to the newest VERSION of the UL standard. It is possible to identify this and model information even if you have likewise thrown away the instructions and the box materials of your multi-pack of GFCI receptacles. They can likewise email you replacement instructions for your devices. IIRC (it was either Lutron or Leviton) there was a recall for a limited production period just as the newest standard requirements were to have been incorporated - but there was a bunch which did not meet the newest version of the standards and had a problem with nusiance tripping that was also not in keeping with the prior standard version.

Mushy, residential snap switches have a long contact spring and when new designed to be "quiet" and can make a (not complete) contact earlier in the switch to on and maintain longer during the switch to off, relying on the spring for contact, idea being to avoid an arc within (but can still arc, and makes less than full electrical contact should the spring be fatigued). Commercial, Industrial non-mushy snap switches have a smaller "noisy" contact and are meant to be operated cleanly and quickly, thereby avoiding an arc when "snapped" fully open or closed but arcing can take place if they are operated slowly or incompletely. They are designed to make full contact or no contact (all or nothing) - where the mushy "quiet" residential grade snap switches rely on the spring metal for contact.

I am bothered that you represented originally you used "hospital grade" switch. What is the rating for the fan - you may not exceed a certain amp rating, VA, H/P rating (inrush) and share the circuit with receptacles. One of the reasons bathroom receptacles are required 20 amp circuits GFCI protected (may have 15 amp face) is the use of personal grooming appliances which oftentimes exceed the safe usage for a circuit protected with 15 amps (i.e. 14 AWG) wiring.

Originally Posted by Stephen DeCosta

The GFCI is LEVITON, [CAT 78999-1] 20 amp rating

I think you might have meant Cat. No. 7899-i (i for ivory)
See (direct link to Leviton's site): 7899-I > 20 Amp > GFCI > Tamper Resistant > All Leviton Products from Leviton Electrical and Electronic Products

Originally Posted by Stephen DeCosta

-- the switch is G.E. co ,U.L. INC, 20 , FED. SPEC. - W-S-896d , with
20 amp rating.

Federal Specification W-S-896d is a specification testing standard not a part/model/catalog number. The (at least used to be) UL test requirements for a switch of this type is about 30,000 cycles under given test conditions set out in (or used to be set out in) UL 20. The switch that passes a more stringent Federal Specification test, W-S-896d, requires an additional 40,000 cold lamp load cycles when combined with the UL 20 test.

Don't know about a GE branded switch SPST snap switch does GE even make these days? To operate the fan you'd have to have an "indicating" type switch so the usual Decora style without ON and OFF markings wouldn't suffice for the motor (fan) control. Curious if you have enough fill room for your box, especially if an older retro-fit. Are you using some sort of wireless switch or timer switch, dimmer or speed control switch to operate the fan?

Many times a fault condition can be created when replacing the items into the box. That's why the old timers advocated only using side screw terminals, always using an insulated ground jumper, taping around the sides of the switch terminals (all around 2-1/2 times) and careful "accordian" or "fan" folding the conductors back into the box with the switch and/or receptacle. Its also why us "old timers" pre-twisted (and even older pre-soldered or tinned) before we applied wire nuts (and we never used the cheap unlabeled supplied wire nuts that came with devices - rarely did they supply a real 'greenie' either).

Originally Posted by Stephen DeCosta

The trip never happens when turning on fan. nor any other situation except occasionally when shutting fan off.

Condensation or other true fault conditon could be of issue. Off (open) should be down - that's also where in the box moisture would first collect and accumulate (gravity - plus cooler temps lurking lower in the box).

I want to thank everyone for all their input. It is very educational.
The switch was given to me from a hospital renovation, maybe using used parts is not the best option. It is "off" in the down position , and the 12 gauge wire to it, feeds on a down slope so if any moisture does enter , it will drain. The fan is up high near the ceiling, and about 5 feet away from a shower entrance.
Since there is no safety issue by wiring to the line side , in this case, when I get the time
that may be the easier way out .
If during an inspection , I find a GFCI that trips with the button , tests good on wiring,
but does not trip with tester , I will identify the issue and suggest to client to have it checked by a licensed electrician for possible inadequate ground.

Originally Posted by Stephen DeCosta

If during an inspection , I find a GFCI that trips with the button , tests good on wiring,
but does not trip with tester , I will identify the issue and suggest to client to have it checked by a licensed electrician for possible inadequate ground.

I hope you know that the GFI does not need a ground in order to operate properly.

Originally Posted by Jim Port

I hope you know that the GFI does not need a ground in order to operate properly.

Jim

Yes he does...now...this was mentioned in the first part of the thread. I think what Steve is saying is that in the future when using his tester and he can't get the gfci to trip he will suspect an open ground. I'm not sure in this case if his tester was showing good ground but he found a bad splice at the panel using no mechanical connector.... repaired and the tester tripped the gfci.

EDIT: Should mention that the 65-501 ideal tester has had some recalls but it uses the hot wire and egc to create a leak current of 8 ma or so to trip the gfci . Etcon does the same. If the test button on the gfci trips the gfci then the gfci is considered operational. I usually never use the gfci testers to trip the gfci itself. I use the test button on the device. I use the trip button on a tester like the etcon only on protected receptacles.

I use the trip button on a tester like the etcon only on protected receptacles.
Thank you guys ,
Roger , on protected receptacles,is there more than one way of defining
the physical nature of what that implies?

Originally Posted by Stephen DeCosta

I use the trip button on a tester like the etcon only on protected receptacles.
Thank you guys ,
Roger , on protected receptacles,is there more than one way of defining
the physical nature of what that implies?

Sure Steve

Sorry if I wasn't clear. Regular receptacles can be wired to the load side of a gfci receptacle and inherit its protection or become gfci protected. You plug your tester in to these receptacles...which should be labeled that they are protected receptacles...then push your test button... the gfci that protects those receptacles should trip. Sometimes you will have to be familiar with where a receptacle should be gfci protected. For instance it is not uncommon for a garage receptacle to be gfci protected from a bathroom gfci in an older home before bathrooms were required to have there own gfci branch circuit. Or if there are multiple receptacles in a bathroom or bathrooms and a gfci receptacle is used the other receptacles will likely be protected by that one gfci so you plug your tester into those to see if the protecting gfci trips.

Here is a drawing to show what I mean

Thank you Roger , that does make it clear to me.