Now I believe

Status
Not open for further replies.
ptonsparky

ptonsparky

Tom
Location
NE (9.06 miles @5.9 Degrees from Winged Horses)
Occupation
EC - retired
I didn't get to wire my new house.

We lost two small freezer loads due to a gfci receptacle tripping in the last year. New freezer. Recept that trips is on the end of the circuit that also feeds the utility/bathroom lights and vent fan. No load goes out of the box the recept is in. I finally got around to putting in a different brand GFCI, Leviton, and checking the freezer. Megger and the whole bit. Perfect. OK, chalk it up to a bum gfci. Leaving the little room this evening I reach up to shut the fan & light off. CLICK!! They are NOT on the load side of the GFCI. At least this one is loud enough to hear when it trips. Reset it & repeated the procedure. CLICK!! Narrowed it down to when the vent fan is shut off the GFCI trips.

Thirty years and I have never had a load that was not on the GFCI trip one. Don't really want to break out the 43B just for a bath fan. I will replace the switch but...If someone has a suggestion I'll take it. Don't even want to say what the wife thinks I should do.
 
kwired

kwired

Electron manager
Location
NE Nebraska
I have one in my shop that trips randomly when shutting off an 8 foot 2 lamp T8 fluorescent that is not connected to the load side of the GFCI. It took a while to figure out what was causing the tripping. The GFCI will trip with or without load on it, you can not make it trip when you want it to, it is very random but always at same instant you turn off the light.

The only thing I usually lose power to if I do not notice it is engine heater on my truck when it is cold outside. I don't think it is cold related as this receptacle has had random tripping for a long time in warm or cold weather. I just more recently noticed it trip when turning off the light and have tried to pay closer attention and noticed when it trips is always when turning off the light.

Last winter when I had to run a generator because of power loss I also had several GFCI trip when the generator ran out of fuel and it puzzled me. I happened to be in my kitchen I heard the generator going down and while it was still on its way down a heard a "pop" in the house. After getting powered up again I realized the "pop" was every GFCI in the kitchen tripping at the same time. went around the house and checked other GFCI's and most of them were tripped did not seem to matter if there was load connected or not but not all of them were tripped.
 
ptonsparky

ptonsparky

Tom
Location
NE (9.06 miles @5.9 Degrees from Winged Horses)
Occupation
EC - retired
Wife wants the freezer in this room. In a main floor utility room that has a stool and utility sink. I would have classified it as a bathroom with a washer and dryer in it. Not much any of us can do about a fan motor that can trip a gfci in this manner. Other than not use that mfg.

Makes me wonder if snubber would help here. I will break out the toys and play the next rainy Saturday we have.
 
Rockyd

Rockyd

Senior Member
Location
Nevada
Occupation
Retired after 40 years as an electrician.
Come on Bob! He came here because this is where you find GOOD electrical contractors, electricians, etc. :)
 
G

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
100405-0910 EST

ptonsparky:

With my normal simple uncontrolled tests on a Leviton 7899 I have not been able to cause it to trip with the switched load on either the input or output side of the Leviton. This was with an 8 ft Slimline or a 175 VA transformer. Then I tried an old AB #2 starter. On the input side of the Leviton it did not trip. But on the load side it did trip. So this invalidates any previous statements I have made that I could not transient trip the Leviton. If I tried more times it would probably trip from the input side.

Looking at the internal electronic circuit it looks like the Leviton is well protected from transient triggering. But from your experience and my above test it is not sufficiently protected. Internally Levition does use an MOV across the line. The question is how is the transient getting to the National chip or the SCR?

I added a GE V150LA10 MOV to the output terminals and for a reasonable number of cycles I did not trip the GFCI using the #2 starter. I also separately tried a 2.25 mfd oil filled capacitor across the output terminals and this alone did not reduce the probability of tripping.

Using a randomly switched inductive load does not provide a well controlled experiment, but it is useful to prove that failure is possible.

Much more controlled experiments should be performed. Then, Leviton and others need prodding to reduce their transient susceptibility.

.
 
charlie b

charlie b

Moderator
Staff member
Location
Lockport, IL
Occupation
Retired Electrical Engineer
So, what exactly is this room, and why does it have a GFCI in the first place? As I read the article 100 definition of "bathroom," I don't see the phrase "utility sink." I see the word, "basin," and my dictionary defines that as, "a SHALLOW (my emphasis), usually round container or pan, used chielfy to hold water or other liquid, esp. for washing." Ok, yes, you could wash your hands in a utility sink, but that does not cause it to become a "basin." Absent a "basin," you don't have a "bathroom," and you don't need GFCI.
 
kwired

kwired

Electron manager
Location
NE Nebraska
charlie b said:
So, what exactly is this room, and why does it have a GFCI in the first place? As I read the article 100 definition of "bathroom," I don't see the phrase "utility sink." I see the word, "basin," and my dictionary defines that as, "a SHALLOW (my emphasis), usually round container or pan, used chielfy to hold water or other liquid, esp. for washing." Ok, yes, you could wash your hands in a utility sink, but that does not cause it to become a "basin." Absent a "basin," you don't have a "bathroom," and you don't need GFCI.
Click to expand...

Does it matter? The fact is there is a GFCI and is probably required for the receptacle outlet but not the lighting outlet(s). In the OP case as well as the one I presented the GFCI trips because of something happening on the line side of the GFCI.
 
Buck Parrish

Buck Parrish

Senior Member
Location
NC & IN
I was at the big blue store the other day and I noticed a lot of freezers come standard with alarms built it.
It's new but thier coming.
 
ptonsparky

ptonsparky

Tom
Location
NE (9.06 miles @5.9 Degrees from Winged Horses)
Occupation
EC - retired
This freezer sits adjacent to the " not a basin". I have already had more than one GC sit on it with their feet in the water. Whatever you want to call the room is purely academic. The problem is with the gfci. The reason for switch replacement is to eliminate any noise generated by it during its action.
 
ELA

ELA

Senior Member
Occupation
Electrical Test Engineer
Similar GFCI "fan shut off" tripping post

Similar GFCI "fan shut off" tripping post

http://forums.mikeholt.com/showpost.php?p=1123858&postcount=22

Please see this similar discussion.

The problem may be that the MOV in the GFCI clamps the transient from the fan but does not limit the rise time (RF content).

Your thought on adding a snubber could help or the EMI filter should do the trick.
 
G

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
100405-2026 EST

ELA:

My guess also is that the source is dv/dt as you are suggesting. Looking again at the board there are long leads to the SCR. There is a capacitor C10 from anode to cathode but not next to the SCR. Then the shunt capacitor C2 from gate to cathode of the SCR is at the 1851 pins and then long leads to the SCR. Probably this capacitor should be close to the SCR pins and a series resistor close to the C2 and then the long lead to the 1851 output.

I have not gone back to try to trigger with the #2 starter from the input side. The Leviton MOV appears about the physical size of the GE I put on the output. I can not trace the part number to a data sheet. This Leviton part number appears to be MU503 and the manufacturer may be MDC.

I believe the GE I used on the output clamps at about 400 V at 1 A.

Between input and output there is some inductance, more than just the leads, due to the current transformer.

Do you have access to any test equipment to perform more controlled experiments than random switching of an inductive load?

.
 
G

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
100405-2111 EST

Looking at the board C10 and C2 can be moved to be close to the SCR, and no problem to add a resistor. Does not appear to be a space problem.

The easiest experiment is to add a cap across the gate to cathode, maybe 0.1 mfd, leave the current cap C2 where it is, cut the trace to the 1851, and add a 100 ohm resistor.

.
 
Status
Not open for further replies.
Top