Tripping Factory 208v cord GFCI on Air Cond.

Status
Not open for further replies.
S

sevlander

Member
Location
new jersey
An existing, functioning window air conditioner(fairly new) with built into plug mounted GFCI trips after a new lighting install in a school. The building and panels are approximately 45 years old and utilize the conduit for the grounding. The existing Romex feed to A/C unit was upgraded to MC and worked for weeks.The A/C is fed on it's own 20 amp 2pole breaker and the lights have their own 1pole breakers. The existing lighting was upgraded shortly after on existing EMT 120v circuits. The new lighting is 2 circuits. One consists of 6 2X4 t-8 fluoresents with electronic ballasts. The other is 10 hi hats with 75 watt incandesant lamps on a 3-way switch and a 3-way dimmer. When the hi hat is turned on the A/C GFCI trips. All circuits originate in the same panel. Can anyone explain or suggest how to solve or troubleshoot this problem? Thanks
 
I might start by meggerring the internal components of the AC unit. There may be a good reason that the molded LCDI is tripping, and it may have nothing to do with your lighting upgrade project. Just coincidently happened at the same time.
 
coincidence?

coincidence?

Thanks, but it ONLY trips when the lighting is turned on. Never when running by itself. Wouldn't you think if it was the A/C unit it would trip at some other point also? I believe it has to be some kind of surge or spike through the ground caused possibly by the electronic dimmer. Some kind of groundind issue?You could'nt meg through the cord could you without possible damage? Are you suggesting megging internal parts separately? If a120volt GFI monitors current imbalance through the neutral, how does a 208volt GFI work?
 
sevlander said:
If a120volt GFI monitors current imbalance through the neutral, how does a 208volt GFI work?
Click to expand...
All GFCI's work by sensing imbalance on all current-carrying conductors. If there is a neutral, it's included. The EGC is never included.

The only requirement of the supplying system is that it has a grounded conductor. Your 208v AC is supplied by two ungrounded conductors.

Logically, the only way the lighting could affect the plug's GFCI is if one conductor of the lighting was connected to a conductor in the AC's power cord.
 
That is how I figured it to work, but the problem never existed before the new lighting was turned on. No interconnection exists other than being out of the same panel. So the mystery is weather a new fault is occurring in the A/C unit or possibly the bult in GFCI in the plug is failing. What testing can show the problem?
 
sevlander said:
What testing can show the problem?
Click to expand...
Substitution, which in the case of an AC unit, is not easily done.

How about temporarily replacing the 3-way dimmer with a 3-way switch?
 
You could temporarily install an EMI filter at the service on the lighting circuit to filter any noise from the electronic ballasts.
If that solved the problem it could be likely that the GFCI built into the air conditioner plug is overly sensitive to EMI.

If required,
would it be possible to cut off the built-in plug/GFCI and replace it with a regular plug? Then install a better quality GFCI breaker at the service?
 
LCDI not GFCI

LCDI not GFCI

I researched and found out it is an LCDI, not a GFCI. An AFCI or LCDI (leakage current detection interupter) is now required on cord connected air conditioners. Supposedly if the cord is pinched or damaged it trips according to the A/C unit's instructions. The problem is the cord appears fine and the tripping still only happens when the 3 way dimmer is turned on. We are going to replace dimmer with a regular 3 way switch and see what happens. I'm not sure how to read or meg the cord on the A/C unit. Any other suggestions?
 
ELA said:
You could temporarily install an EMI filter at the service on the lighting circuit to filter any noise from the electronic ballasts.
If that solved the problem it could be likely that the LCDI built into the air conditioner plug is overly sensitive to EMI.

If required,
would it be possible to replace the built-in plug/LCDI and replace it with a regular plug? Then install a better quality GFCI breaker at the service?
Click to expand...

Then take the removed LCDI and cord you removed and connect a (220v) lamp load to it. Plug it in and turn on the lighting to see if it then trips.
This will eliminate any questions about the Air Conditioner load as the issue vs. an overly sensitive current sensing device.

I realize you may not be able to permanently replace the attached cord. You had asked for troubleshooting ideas. You could also temporarily wire in an EMI filter just ahead of the Air Conditioner cord, however it would be better at the service or at the ballasts (if they were the noise source).

Contrary to what some people contend some GFCI's will trip due to noise and harmonics. Some are more immune than others.
 
gfci tripping

gfci tripping

Some newer electronic devices we install use the ground wire as the return for high impedance electronics. I suspect the 3 way dimmer is the culprit as others have suggested lose the dimmer and I think your problems are solved. If you still want to use the dimmer pull ground wires back to the panel to eliminate any stray current running the wrong way due to mechanical grounding problems in the pipe system. I have had 30 yrs of watching dirtbags install wiring that they know is deficient and Jimmy Crack Corn and I dont care attitude. I dont get shocked by the disreguard for safety anymore I just wish for accountability for those dirtbags to pay when someone gets hurt.
 
Last edited:
ELA said:
Contrary to what some people contend some GFCI's will trip due to noise and harmonics. Some are more immune than others.
Click to expand...

Sure would like to see some proof of that, IMO a tripping GFCI is working and the circuit has an issue.

A GFCI is a fairly basic device.
 
I cannot site any references but I do have direct experience. I spent may years working on electrical noise issues. There are many instances of electronic devices being upset by electrical noise.
For example most equipment incorporating integrated circuits can be upset if enough electrical noise is injected into it. It does not necessarily need to be a large voltage if the frequency is high enough. Often time relatively small amounts of common mode noise at a high frequency can cause a microprocessor to reset (computers being one example).

I have also many times experienced electronics resetting when exposed to noise even though the devices operated on a "differential principle". Such as the differential current required to trip a GFCI. If enough common mode noise is injected it can cause the electronics to reset even though there is no measureable "difference" in current.

I once worked on a robotics system that an engineer decided should be protected by a GFI (30 ma trip). Once in a while , and very sporatically the GFI would trip. This would only occur when the servo drive was decelerating.
The GFI would never trip on inrush (a truely differential event) or when idle or when moving slowly. It would only trip when the motor was in a heavy deceleration (regeneration occuring). I replaced the GFI with several different units and types and the problem persisted.

In this particular case we eventually just removed the GFI since it was not really required by any stanards. The engineer just thought it would be a good idea.

I no longer work for the company where I performed EMC testing. So I no longer have access to the type of equipment required to demonstrate how noise can affect electronics in ways that do not to follow the normal expectations.

If I still worked there I would be happy to perform a test and present the data.

The two current posts here with GFCI issues tripping both use electronic ballasts which are notorious for their electrical noise output. I know it may be difficult to perform diagnostic testing at a customer site but here are two examples where when the noise source is removed the problem goes away.

I hope that another answer can be found but if not then consider filtering the noise source or eliminating it.
 
iwire said:
Sure would like to see some proof of that, IMO a tripping GFCI is working and the circuit has an issue.

A GFCI is a fairly basic device.
Click to expand...

Never having searched for information outside of my own experience I just did. You can find a fair amount of information on how GFCIs should be tested for immunity to electrical noise.

Here is a link to a Leviton device. See note #4 there.

http://www.bayneselectric.com/html/pressgfci.html

The better quality manufacturers will include a specification as to how much noise the device can withstand. It appears to me that the manufacturers of these devices are well aware that they can false trip due to noise.

It has been my experience that some are better designed than others. In one instance I changed a Seimans earth leakage breaker with another manufacturers (used in an electrically noisy environment) and the nuisance tripping problem disappeared. The other manufacturer specified its noise immunity levels whereas the Seimans product did not.
 
Status
Not open for further replies.
Top