Have installed a comb. vent fan and heater unit in bathroom.Fed from load side of GFCI has three switches in circuit Lav light, vent fan, heater when the customer turns the vent fan switch to the OFF position trips GFCI recept. Doesn't trip everytime. any help would be appriciated.
GFCI and Vent Fan
- Thread starter Raccoon 3
- Start date
- Status
rcarroll
Senior Member
- Location
- Fort Collins, Colorado
Is the fan/heat combo installed outside of the tub or shower space? If so, check the instructions of the unit. It may not have to be GFI protected at all.
I realize this doesn't answer your question, but it may solve the problem.
I realize this doesn't answer your question, but it may solve the problem.
1793
Senior Member
- Location
- Louisville, Kentucky
- Occupation
- Inspector
I had this same issue a while back and I did solve it, I'll have to look back and see if I made notes as to how I did it. I think I opened all of the connections and made sure all were tight and correct.
I feel you pain.
I feel you pain.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
091039-1721 EST
It is likely caused by a voltage transient from turning off an inductive load, the fan motor. However I would not expect a small fan motor to produce that large a transient.
Is it more likely to occur when nothing else is on, and the fan is turned off?
Is it less likely to occur when the heater is on and remains on when the fan is turned off?
Identify the year of manufacture and brand of the GFCI.
I ran experiments on a recent Leviton GFCI to see if my test would trip the GFCI. It did not trip. The test was rapidly turning on and off an 8 ft Slimline with a magnetic ballast and no line filter to generate line transients. The rapid part is only to more quickly generate a large transient in a short time.
.
It is likely caused by a voltage transient from turning off an inductive load, the fan motor. However I would not expect a small fan motor to produce that large a transient.
Is it more likely to occur when nothing else is on, and the fan is turned off?
Is it less likely to occur when the heater is on and remains on when the fan is turned off?
Identify the year of manufacture and brand of the GFCI.
I ran experiments on a recent Leviton GFCI to see if my test would trip the GFCI. It did not trip. The test was rapidly turning on and off an 8 ft Slimline with a magnetic ballast and no line filter to generate line transients. The rapid part is only to more quickly generate a large transient in a short time.
.
AV ELECTRIC
Senior Member
Is this a dedicated circuit or coming off the bathroom circuit?
electricmanscott
Senior Member
- Location
- Boston, MA
- Occupation
- Massachusetts Master Electrician, one man show.
Why is this on a GFI to start with?
I would also start by checkiong all connections. I think I had this problem with a Panasonic fan/light. Re-did connections and problem was solved. Maybe a strand of wire peeking out of a wirenut.
I would also start by checkiong all connections. I think I had this problem with a Panasonic fan/light. Re-did connections and problem was solved. Maybe a strand of wire peeking out of a wirenut.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
091031-0628 EST
Raccoon 3:
I ran my transient test again on a new Leviton 7899-KT SmartLockPro and could not produce a false trip. Inductive load is magnetic ballast 8' Slimline.
If all your load wiring comes from the hot and neutral output terminals of the GFCI, then it does not matter how you wire the load or loads relative to the GFCI operation. Obviously you can not have any leakage sneak paths around the GFCI.
electricmanscott's point is important to look into.
Your tripping only on turn off of the fan motor sure seems to imply an inductive kick problem. Have you tried replacing the GFCI with another Leviton? From my analysis of the present Leviton GFCI circuit it should be relatively immune to short voltage transients.
Maybe there is insulation breakdown in the fan motor system such that an inductive kick on turn off would produce a path to ground (EGC). If there was an MOV at the fan motor from hot to EGC, then this could be such a sneak path.
Do you have an easy portable load that would be inductive that could be repetitively plugged into the GFCI receptacle? I tried a Universal 806-SLH-TC-P with no lamps and I could get nice arcs from the wire to the socket and no tripping. No plug just put one wire into the socket and repetitively make and brake contact with the other wire. Don't come into contact with the high voltage lamp wires. Unloaded these are about 420 V.
.
Raccoon 3:
I ran my transient test again on a new Leviton 7899-KT SmartLockPro and could not produce a false trip. Inductive load is magnetic ballast 8' Slimline.
If all your load wiring comes from the hot and neutral output terminals of the GFCI, then it does not matter how you wire the load or loads relative to the GFCI operation. Obviously you can not have any leakage sneak paths around the GFCI.
electricmanscott's point is important to look into.
Your tripping only on turn off of the fan motor sure seems to imply an inductive kick problem. Have you tried replacing the GFCI with another Leviton? From my analysis of the present Leviton GFCI circuit it should be relatively immune to short voltage transients.
Maybe there is insulation breakdown in the fan motor system such that an inductive kick on turn off would produce a path to ground (EGC). If there was an MOV at the fan motor from hot to EGC, then this could be such a sneak path.
Do you have an easy portable load that would be inductive that could be repetitively plugged into the GFCI receptacle? I tried a Universal 806-SLH-TC-P with no lamps and I could get nice arcs from the wire to the socket and no tripping. No plug just put one wire into the socket and repetitively make and brake contact with the other wire. Don't come into contact with the high voltage lamp wires. Unloaded these are about 420 V.
.
76nemo
Senior Member
- Location
- Ogdensburg, NY
- Location
- Massachusetts
Yeah, where is the GFCI device in relation to the shared neutral?
Power Tech
Senior Member
- Location
- Lebec, California
Like $ said, why are you putting this on a GFI?
Personally I run a separate circuit for the H/L/F.
Another 20 amp just for the GFI.
IMO you are going to get call backs the first hair dryer that is plugged in.
Personally I run a separate circuit for the H/L/F.
Another 20 amp just for the GFI.
IMO you are going to get call backs the first hair dryer that is plugged in.
greenjeans
Inactive, Email Never Verified
- Location
- Now living in Virgina
share neatral, no good if heaters on gfi and lights or not. i would put it all on one circuit and gfci (if needed) or removed gfci altogether. IMO
mcclary's electrical
Senior Member
- Location
- VA
My thoughts exactly, and that's the way I do it. I did a bathroom that had two. I brought each one it's own circuit. I never come off the GFI. I run the fan off the 15 amp lighting circuit, and bring a #12 to the heat.
motorkitty
Member
Which H/F/L combo did you buy? There are very few of them that you can get that don't[I need a dedicated circuit. If it is a Panasonic FV11VHL2 then you have to have a dedicated circuit. And at minimum 20 amp. As far as your GFCI breaker, (make sure the amp rating is high enough) if you know it is not faulty, and it continues to trip, then there is a short stemming usually from the installation of the unit.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
091103-1319 EST
I agree that if a GFCI is not required, then do not use one. When not necessary it is just one more complex item to cause problems.
However, I would like to know the cause of tripping of the GFCI of the first post (OP). When conditions exist that cause a GFCI to randomly trip, or trip for a known trigger (in this case turning off the fan), but a condition for which the GFCI is not expected to trip, then it is desirable to try to determine why the GFCI is tripping. Is it an individual faulty GCFI, a design problem, or an unusual event?
In the case of this model GFCI I would not expect it to trip on typical voltage transients I have generated. And I do not think the fan motor turn-off is likely to generate transients as large as with my fluorescent ballast test. Is the particular GFCI faulty? Substituting a different GFCI (meaning a different batch of the same model or a different design) would be a typical first test. Second try bench testing the particular GFCI to see if the tripping can be caused in a different environment.
I did run a test with an MOV from the output hot to ground and cycled the ballast and there was not enough transient energy to ground to trip the Leviton GFCI. As an aside the leakage current of the GE V150LA10A MOV was about 0.05 MA with 124 V RMS 60 Hz applied. This MOV is about 1/2" in diameter. Maximum specified peak transient clamping voltage at 10 A is 340 V, minimum DC voltage at 1 MA is 184 V.
Why is it important to know why a failure occurs? Because it can determine your decision of what products to use, or how to help in the improvement of the products.
.
I agree that if a GFCI is not required, then do not use one. When not necessary it is just one more complex item to cause problems.
However, I would like to know the cause of tripping of the GFCI of the first post (OP). When conditions exist that cause a GFCI to randomly trip, or trip for a known trigger (in this case turning off the fan), but a condition for which the GFCI is not expected to trip, then it is desirable to try to determine why the GFCI is tripping. Is it an individual faulty GCFI, a design problem, or an unusual event?
In the case of this model GFCI I would not expect it to trip on typical voltage transients I have generated. And I do not think the fan motor turn-off is likely to generate transients as large as with my fluorescent ballast test. Is the particular GFCI faulty? Substituting a different GFCI (meaning a different batch of the same model or a different design) would be a typical first test. Second try bench testing the particular GFCI to see if the tripping can be caused in a different environment.
I did run a test with an MOV from the output hot to ground and cycled the ballast and there was not enough transient energy to ground to trip the Leviton GFCI. As an aside the leakage current of the GE V150LA10A MOV was about 0.05 MA with 124 V RMS 60 Hz applied. This MOV is about 1/2" in diameter. Maximum specified peak transient clamping voltage at 10 A is 340 V, minimum DC voltage at 1 MA is 184 V.
Why is it important to know why a failure occurs? Because it can determine your decision of what products to use, or how to help in the improvement of the products.
.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
091003-1947 EST
Another rough experiment.
Leviton 7899 GFCI, DeWalt radial arm saw. Approximate several second inrush about 70 A on start-up. First half cycle on some cycles will be considerably higher.
No tripping of the 7899 over a period of about 6 on-off trial cycles.
.
Another rough experiment.
Leviton 7899 GFCI, DeWalt radial arm saw. Approximate several second inrush about 70 A on start-up. First half cycle on some cycles will be considerably higher.
No tripping of the 7899 over a period of about 6 on-off trial cycles.
.
12-3
12-3
Copy that. 210.(C)(3)Bathroom Branch Circuits. In short, 20A BC shall be provided to supply bathroom receptacle outlet(s). etc....Such circuits shall have no other outlets. Sharing a neutral would be non-compliant to this...I believe. rbj
12-3
Copy that. 210.(C)(3)Bathroom Branch Circuits. In short, 20A BC shall be provided to supply bathroom receptacle outlet(s). etc....Such circuits shall have no other outlets. Sharing a neutral would be non-compliant to this...I believe. rbj
ELA
Senior Member
- Occupation
- Electrical Test Engineer
Gar,
Have you ever measured the transient levels generated by your test ballast?
When talking about Electric Fast Transients what is important is "knowing" the transient generation levels.
Unfortunately I have never been able to locate any GFCI manufacturers specification as to their immunity levels against EFT(referenced to a standard).
If you knew what the manufacturer's claimed immunity level was, and you had a repeatable -calibrated EFT generator- then you could make some meaningful comparisons of GFCI devices.
Another cheaper type of EFT generator is referred to as the NEMA Chattering Relay Test. I have built these in the past but they are like your ballast test in terms of being uncalibrated - likely more repeatable though.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
091104-0620 EST
ELA:
In the past under some arbitrary conditions several thousand volts. But very dependent upon on the circuit and where measured. I can not afford any more sophisticated equipment than my #2 motor starter or the fluorescent. Thus, uncontrolled experiments, but somewhat representing possible sources in my environment.
An aside. Most of my 8' Slimline fixtures have an input line filter that I installed to reduce EMI to radios. Recently when I started playing with the Kill-A-Watt instrument I was surprised to see the great effect the line filters had on the power factor. These filters only have small capacitors, and some series inductance.
From what I see in the Leviton 7899 circuit it should be relatively immune to directly coupled voltage transients of the levels I might expect in its typical use.
Thus, it is interesting that several threads are popping up presently about false triggering of GFCIs. Are these really false triggers, or are the GFCIs really indicating a true problem?
.
ELA:
In the past under some arbitrary conditions several thousand volts. But very dependent upon on the circuit and where measured. I can not afford any more sophisticated equipment than my #2 motor starter or the fluorescent. Thus, uncontrolled experiments, but somewhat representing possible sources in my environment.
An aside. Most of my 8' Slimline fixtures have an input line filter that I installed to reduce EMI to radios. Recently when I started playing with the Kill-A-Watt instrument I was surprised to see the great effect the line filters had on the power factor. These filters only have small capacitors, and some series inductance.
From what I see in the Leviton 7899 circuit it should be relatively immune to directly coupled voltage transients of the levels I might expect in its typical use.
Thus, it is interesting that several threads are popping up presently about false triggering of GFCIs. Are these really false triggers, or are the GFCIs really indicating a true problem?
.
ELA
Senior Member
- Occupation
- Electrical Test Engineer
The EFT testing is designed to simulate arcing contacts when switching an inductive load. If you might be interested in reading about EN61000-4-4 and EFT testing you can see that the goal is to produce multiple transients in close proximity to each other:
http://www.reo.co.uk/en_61000-4-4_electrical_fast_transients
The PDF download is informative.
While an MOV will limit the maximum amplitude of the incident transient it does not remove the high frequency components of the fast rise time wavefront. That is why it is important in any good EMI protection scheme to follow the MOV with some sort of LC filtering.
When I ran calibrated EFT tests in the past I would often find it difficult to upset a piece of equipment when running the generator in a "single shot" mode where only one +/-2KV transient was produced.
When the generator was allowed to free run and thus producing "bursts" of transients the equipment could often be readily upset.
In these recent posts involving turning off a fan it certainly does sound like nuisance tripping due to the transients produced when the fan is turned off.
The key to replicating these circumstances may be both the inductive charactoristics of the load plus the "bounce" of the switch that interrupts the load.
http://www.reo.co.uk/en_61000-4-4_electrical_fast_transients
The PDF download is informative.
While an MOV will limit the maximum amplitude of the incident transient it does not remove the high frequency components of the fast rise time wavefront. That is why it is important in any good EMI protection scheme to follow the MOV with some sort of LC filtering.
When I ran calibrated EFT tests in the past I would often find it difficult to upset a piece of equipment when running the generator in a "single shot" mode where only one +/-2KV transient was produced.
When the generator was allowed to free run and thus producing "bursts" of transients the equipment could often be readily upset.
In these recent posts involving turning off a fan it certainly does sound like nuisance tripping due to the transients produced when the fan is turned off.
The key to replicating these circumstances may be both the inductive charactoristics of the load plus the "bounce" of the switch that interrupts the load.
- Status