New GFI 110v outlets blow out TECO VFD

[mksj]

There is no EMI/RFI filtering on the Teco VFD unless you add one. There is extensive information as too the leakage issue with VFDs, some manufactures have taken extensive and costly measures to minimize this and electrical noise radiation, but not at this level. The design of the line filter has an effect on the GFCI trip characteristics, the ones I typically use for electrical noise attenuation will outright trip them because of a high (intentional) leakage current to ground. I have used the TDK/LAMBDA-MC1230 noise filters on some VFD systems, they are rated at less than 1 mA of leakage current. Worth a try, connect it close to the VFD.
https://www.ebay.com/itm/LAMBDA-MC1230-Surplus-New-In-factory-packaging/391809295708

Sometimes there can be slight variations between GFCI's as to tripping, but they all most meet the same trip criteria for a Class A ground fault. I have worked with a number of individuals who have used the same Teco and all had the same issue with the GFCI tripping (these were the 120V version). There may be some variations from VFD model to model brand to brand, but this is a "known" problem with VFDs with no quick or "cheap" fix other than to wire a socket without a GFCI. You will have one GFCI feeding a branch of downstream sockets, you can move the GFCI to the next socket down and use the first as a non-GFCI and if you need to meet code, hard wire it. QED.

ABB Reference document. These have already been stated.
LINE SUPPLY GFCI CIRCUIT BREAKER’S AND AC DRIVES

Description:
Can a GFCI circuit breaker be installed on the AC power feeding a drive, or will the drive cause it to trip?

Answer:
The AC Drive most likely will cause the GFCI protection device to trip. The reason the AC Drive will cause this tripping of the GFCI is the Common Mode Current or Common Mode Noise (CM Noise) that the VFD is producing. The high rate that the IGBT is switched on and off is around (1 - 16 kHz). This switching creates Common Mode Electrical Noise. The Common Mode Noise is the current that appears on a conductor with respect to ground. This Common Mode Noise can and will travel throughout the plants ground system and even beyond. This Common Mode Noise can affect the operation of the application, and other equipment in the plant by interfering with signals from sensors, computers, PLC’s, telephone and radio. The VFD does provide ground fault protection on the output of the VFD.

Documents or other reference material: None

Corrective Actions:
The goal is to have all the Common Mode Noise return to the VFD. Here are some actions that will help reduce or eliminate the tripping of the GFCI. The motor cable should be VFD rated motor cable and the motor and shield should be grounded back to the VFD. The User Manual will provide information on how to correctly install the VFD. The installation of RFI/EMC filters before or inside the VFD. Installing an isolation transformer before the VFD and after the GFCI circuit breaker will help reduce the tripping of the GFCI.

 

[4GSR]

Glen,

Did you by chance ask the inspector about needing the circuit for running a VFD from? He might change his tune and offer a recommendation.

Worse yet, blow his mind up trying to figure out why you have a VFD in a residential setting.

 

[Blackjackjacques]

The GFI is probably sensing a 'fault current' because of the RF filtering in the VFD, but the GFI trip point
is set at (5 mA?) and it is common practice to steal a little current (like, one-third of the trip current)
in some appliances. For RF filtering purposes, a small capacitor on the line/ground will take
acceptable 60 Hz current, below the GFI trip point, BUT will also pass some switching current
at high frequency. Other appliances ditto pass small currents.

So, the VFD, and the load on the VFD, and the filter detailed characteristics, and the capacitance from line
to ground of everything ELSE on the circuit, and the high frequency sensitivity of the particular GFI module,
all combine to determine whether it trips. The production year/month of a VFD might
determine whether the RF filters put 1.5 mA to ground, or 3.5 mA. Perfectly good devices,
no shock or fire safety problems at all, can be incompatible with the GFI.

This suggests that a breaker-box GFI (one unit protecting a half dozen outlets) is more likely
to trip than an end-node GFI (one unit protecting two outlets). It also suggests that a single
component (the powerline filter) of the VFD can be the only problem, and those can be replaced
fairly easily. Me, i'd see if it was possible to clamp a ferrite common mode choke on the
line/neutral wires inside the VFD. Or, ask the manufacturer of the VFD.

GFI is probably sensing a 'fault current' because of the RF filtering
in the VFD, but the GFI trip pointis set at (5 mA?)
.

You mean leakage current and not fault current - don't you? You can feed a line to neutral fault current through a GFCI device all day
until the branch OCP device trips. Also, there is no "set" point for GFCI tripping. GFCI devices follow an inverse time-current curve described by the equation T=(20/I)^1.43, where T is time in seconds, and I is current in mA. So at 5 mA, the GFCI could take over 7.2 seconds to trip, while 20 milliseconds @ 300 mA current. This curve is based on the IEC 60479 time-current exposure curves.

, the VFD, and the load on the VFD, and the filter detailed characteristics, and the capacitance from line to ground of everything ELSE on the circuit, and the high frequency sensitivity of the particular GFI module,
all combine to determine whether it trips. The production year/month of a VFD might
determine whether the RF filters put 1.5 mA to ground, or 3.5 mA.

GFCI devices do not trip at 3.5 mA and below - or at least they are not supposed to and cause for rejection under the UL 943.

suggests that a breaker-box GFI (one unit protecting a half dozen outlets) is more likely
to trip than an end-node GFI (one unit protecting two outlets).

Not true - the breaker style GFCI devices are built to the exact same UL 943 standard as are the receptacle devices, as are the temporary power tap GFCI devices.

It also suggests that a single component (the powerline filter) of the VFD can be the only problem, and those can be replaced fairly easily. Me, i'd see if it was possible to clamp a ferrite common mode choke on the
line/neutral wires inside the VFD. Or, ask the manufacturer of the VFD.

I don't know how anyone be so certain that the VFD "can be only problem" especially based on the information given here, and I also do not believe they can be "replaced fairly easily" As mksj posted, there is no RF bypassing on these devices unless installed after manufacture, so I don't see where clamping a ferrite choke would do any good. With the GFCI a mandatory code compliance requirement, and with the given VFD and load device, the only real code-compliant option the OP has is to employ an isolation transformer.

 

[whitmore]

[about a VFD which might draw 3.5 mA current to ground through an RF filter, and RF filter
standards changing over time]

GFCI devices do not trip at 3.5 mA and below - or at least they are not supposed to and cause for rejection under the UL 943.

But the VFD isn't the only item on the circuit, and it's the sum of that and other items (like, the motor that
the VFD drives, worklights, lube pumps,
cellphone chargers and the like) that matters. Six-outlet branch circuits with one GFCI might
trip when the VFD is plugged in, even if the VFD alone is not a problem.

As mksj posted, there is no RF bypassing on these devices unless installed after manufacture

In my years of poking into powered apparatus, I've never seen a switchmode power supply, as a VFD
certainly is, that did not have a number of RF filter elements on its AC inlet. Maybe there's
not a label that proclaims RF filter, but it's there.

 

[Dabbler]

so... without wanting to offend, and recognizing a diversity of views here, I consulted with an active electrical engineer who specializes in custom RF and power circuits. I'm not trying to pull a 'Authority trump card here', but to provide some additional info for consideration...

If putting simple EMF filters doesn't fix the problem, then there must be either leackage or common mode current exceeding the trip threshold. This is an extremely dangerous situation, despite the fact you have been runing the VFD for years. Repair/replacement is the safest approach. Any new modern VFD should be designed with AFCI and GFI in mind even the commercial units. I was working in a commercial shop a few years ago where all the machines were on $$$ GFI circuits.

 

[ch2co]

My (10+year old) Hitachi VFD came with a rather thick and highly detailed manual that has a chapter about how VFDs can put a lot of high frequency trash into the incoming power line. They provided chokes (inductors) that can be placed in the 3 wire VFD input power line to protect other components on the same line from all this “hash”. Your GFI would be one of these sensitive components. I’ve not run into any problems with me system without such protection, however my VFD is hardwired to the mains line without a GFI. None of my other GFI’s in my shop have given me any troubles, but they are on another set of circuit breakers in my mains panel.





with a

 

[rwm]

I'm changing over to steam....
R

 

[Blackjackjacques]

so... without wanting to offend, and recognizing a diversity of views here, I consulted with an active electrical engineer who specializes in custom RF and power circuits. I'm not trying to pull a 'Authority trump card here', but to provide some additional info for consideration...

If putting simple EMF filters doesn't fix the problem, then there must be either leackage or common mode current exceeding the trip threshold. This is an extremely dangerous situation, despite the fact you have been runing the VFD for years. Repair/replacement is the safest approach. Any new modern VFD should be designed with AFCI and GFI in mind even the commercial units. I was working in a commercial shop a few years ago where all the machines were on $$$ GFI circuits.

I would not be offended, and I am an active and licensed electrical engineer specializing in power and control as well as having served on the UL 913 STP for 15+ years now. Inasmuch as I came here to learn to machine stuff, I am instead loitering in the electrical forum rather than use my brand new Rong Fu 45, that still needs for me to install a DRO still in boxes underfeet. I'll rely on the old adage "if all you have is a hammer, everything looks like a nail. However, I have made progress and last night did put together the MachTach RPM kit.

Getting a 2nd or 3rd opinion is good.

Although the leakage is sufficient to trip a GFCI when using a VFD appliance, there is no danger or risk whatsoever. UL will be the first one to tell you that GFCIs should not be used with such equipment, especially commercial equipment exactly because of the very condition described. In fact, UL will even tell you that many residential equipment, including PCs, printers, etc, may trip a GFCI. If a ground fault system is required, they will point you to the other GFCI classes. I have many pieces of equipment that I cannot use on a GFCI protected branch circuit, including UPS, printers, PCs which I just assign to a non-GFCI circuit. That is code compliant here in my county. For circuits with GFCI protection that I must share with equipment that cannot run off a GFCI, I use isolation transformers to feed a common receptacle bank. I have to do this with a bank of vacuum tube test equipment. I also have a pair of old tube Mac 30 amplifiers that I also feed through an ISO transformer. There is no way that electric equipment manufacturers, especially workshop or commercial equipment vendors will make use on an GFCI (or AFCI) a priority performance condition. In this particular case, I would try to persuade the code official, appeal the lower level decision and ask for a written decision and submit to the appeal board, a written explanation of the problem, as well as include whatever documents (including manufacturer's statements) detailing the conflict. I can't imagine any code authority denying such an application for waiver, especially given what appears to be a draconian interpretation of whatever NEC edition they are invoking. If this fails, then the ISO Xfrmr is the next best solution while maintaining code compliance.

 

[whitmore]

If putting simple EMF filters doesn't fix the problem, then there must be either leackage or common mode current exceeding the trip threshold. This is an extremely dangerous situation...

Good, yes, it's about current exceeding the trip threshold. But, one style of filter puts a capacitor-to-ground
in the circuit, and that capacitor (sometimes called an X capacitor, there's safety requirements for 'em)
might be the only 'leakage' current path. That leakage path is just like the GFCI "test" button, NOT
a shock hazard, but the GFCI doesn't know that. Not 'extremely dangerous', but not
useful either, because the filter, or the test button, shuts the GFCI down.

For (a previous employer's) medical equipment, we had a budget of how many such acceptable
capacitors could be used in a complex system, and it was 2. No shock hazard, but simply plugging three
acceptable and safe devices into a branch circuit could fail safety tests, or trip a GFCI.

The kind of filter one wants to use in a VFD is one that does NOT have an X capacitor, because
the high frequencies a VFD generates can put more current through an X capacitor.

 

[markba633csi]

Pain in the %SS . Look at many 220 volt clothes dryer installations. Two hots and a neutral, no ground there. Well what happens if the neutral floats?
You guessed it- shock hazard. Same thing with many welder installs.
Give me a good old ground wire any day, instead of relying on some fancy-shmancy electronics to save me. Fooey! OK rant over.
Mark S.
Authorities are probably reading this: Look! a malingerer, a malcontent- arrest him!