To avoid these ads, REGISTER NOW!

High Frequency GFI's…….. Some good information for you contractors.

To avoid these ads, REGISTER NOW!
C

Codyboy

Well-known member
Joined
Jan 31, 2019
Messages
1,577
Location
S.E. TEXAS
I wonder why it took this long for them to look into it for a solution since it's been an issue since their invention?
 
M

mm08822

Well-known member
Joined
Jan 13, 2012
Messages
5,763
Location
NJ
I searched for the specifics and short of purchasing UL 101 and UL 943 requirements, I found some related info splaining it.

Also note, the HF functionality is an optional supplement to the newly revised UL 943. So not every GFCI mfr may jump on this right away.

UL stated that a device meeting this optional requirement shall be labeled with "HF" or something similar. This would be in addition to the product spec sheet that could just states meeting UL 943. Per a UL email I have, such marking on the device may not be readily visible from the deadfront but labeled elsewhere such as the device yolk.

However, I purchased gfci recepts 2 weeks ago that has no such HF marking on the device but does state HF in its online description.
 

Attachments

  • CPSC-staff-comment-UL-101-Leakage-Current-0223.pdf
    120.1 KB · Views: 2
  • 2022-07-14-UL-STP-101-for-Standard-for-Leakage-Current-for-Utilization-Equipment.pdf
    84 KB · Views: 1
  • Eaton-HF-GFCI_whitepaper-wp003007en.pdf
    218.3 KB · Views: 4
  • National Electrical Code® 2026 Changes to Chapter Two - IECI.pdf
    812.5 KB · Views: 1
M

mm08822

Well-known member
Joined
Jan 13, 2012
Messages
5,763
Location
NJ
Codyboy said:
I wonder why it took this long for them to look into it for a solution since it's been an issue since their invention?
Click to expand...
B/c it has only gotten worse since vfds becoming mainstream pushed by energy standards imposed on appliance mfrs.

My guess is the early "problems" were real problems that needed to be eliminated which drove the need for gfci protection in the first place.

I remember when gfci recepts first came out (in my world) and were $25-40ish and a regular cheap recept was 35 cents. Everybody had their undies in a bunch. That is why bathroom, garage and exterior recepts were protected by one gfci!
 
U

u3b3rg33k

Well-known member
Joined
Dec 18, 2017
Messages
4,037
M

mm08822

Well-known member
Joined
Jan 13, 2012
Messages
5,763
Location
NJ
u3b3rg33k said:
yes they do

BUT
if there's no non-measured path for the leakage current to return then the GFCI doesn't trip.

this is in reference to VFDs specifically.
Click to expand...
I think Wylie means the gfci isn't measuring current on the grounding conductor itself.

The gfci is measuring the difference between power lines - current in and current out. Yes a ground path is needed for the leakage to ultimately get back to the source.
 
U

u3b3rg33k

Well-known member
Joined
Dec 18, 2017
Messages
4,037
mm08822 said:
I think Wylie means the gfci isn't measuring current on the grounding conductor itself.

The gfci is measuring the difference between power lines - current in and current out. Yes a ground path is needed for the leakage to ultimately get back to the source.
Click to expand...
i agree.

I got out my scope and meter, and even with a filter on the VFD's ground, reducing current on the ground below my meter's ability to measure (I even used the 10x side), the GFCI still trips.
If I disconnect the ground wire, it works just fine, even with a load on it. so IMHO, I'd guess that the GFCI's logic is either being confused by the HF noise, or it's just extremely sensitive to it.
 
M

mm08822

Well-known member
Joined
Jan 13, 2012
Messages
5,763
Location
NJ
Chuckster in NJ said:
Here is an article from NJ DCA to code officials about this "new GFI technology"………… Get ready because inspectors are now educated and must enforce this.
Click to expand...
Today I looked at all Leviton GFCI recepts that are listed in their HF category. NONE of them have any "HF" designation on the device. The only mention of HF in the cut sheet is increased technology capability to reduce trips. The ones I purchased last week have no HF markings either.

I chatted with Leviton today and the chatty Patty there was telling me of all the certs it had, but couldn't answer my q about how does one identify that device as HF. After a little more PC response, she referred me to Customer Technical Services. I'll see what I get in my inbox on Monday.

Even the Eaton CBs didn't have anything.

Legrand claimed "appliance friendly" but never got specific. Still no HF marking found either.

(I'm wondering if passing the UL tests are a challenge and this is all just pre-marketing to stroke the public.)
 
M

mm08822

Well-known member
Joined
Jan 13, 2012
Messages
5,763
Location
NJ
u3b3rg33k said:
i agree.

I got out my scope and meter, and even with a filter on the VFD's ground, reducing current on the ground below my meter's ability to measure (I even used the 10x side), the GFCI still trips.
If I disconnect the ground wire, it works just fine, even with a load on it. so IMHO, I'd guess that the GFCI's logic is either being confused by the HF noise, or it's just extremely sensitive to it.
Click to expand...
This is exactly the same application for a 2 wire gfci receptacle on a system with no egc present. The gfci still works.

What goes in (measured) has to equal what goes out (measured) within 5mA +/-1 mA. That gateway is through the toroid in the gfci.

It's not confused logic. It was written improperly years ago not accounting for the higher frequency leakage values having lesser effect on the muscles in the body. It seems like it now analyzes many bands of frequencies and determines the magnitude of the current in each band. Each band's magnitude is weighted to determine an overall 60hz 4-6mA equivalent.

That's what this chart tells me.....
1778901983100.png
 
Last edited:
W

wyliesdiesels

Well-known member
Joined
Aug 14, 2012
Messages
19,968
Location
Modesto, CA
u3b3rg33k said:
yes they do

BUT
if there's no non-measured path for the leakage current to return then the GFCI doesn't trip.

this is in reference to VFDs specifically.
Click to expand...

they do not. a GFCI can be installed without connection to an EGC

the pathway for leakage current has nothing to do with what i said.
 
Last edited:
OP
C

Chuckster in NJ

Well-known member
Joined
Jan 26, 2010
Messages
2,265
Location
Hunterdon County NJ
wyliesdiesels said:
they do not. a GFCI can be installed without connecting to an EGC

the pathway for leakage current has nothing to do with what i said.
Click to expand...
A GFCI is like a "balance beam" scale…….. What it sees (amps) on the hot wire (3.289 amps) it must see on the neutral wire (3.289 amps) otherwise it will trip. Take a wiggy voltage tester and go from hot to ground and it will trip because there will be a imbalance.
A knob and tube wiring system has no ground and if you replace a receptacle on a knob and tube system it should be a GFI receptacle to be safe.…….. AGAIN, the GFCI does NOT rely on a ground wire for proper operation.
 
To avoid these ads, REGISTER NOW!
R

rlitman

Well-known member
Joined
Oct 18, 2010
Messages
24,553
Location
Long Island
mm08822 said:
...It's not confused logic. It was written improperly years ago not accounting for the higher frequency leakage values having lesser effect on the muscles in the body. It seems like it now analyzes many bands of frequencies and determines the magnitude of the current in each band. Each band's magnitude is weighted to determine an overall 60hz 4-6mA equivalent.

That's what this chart tells me.....
Click to expand...
Ok, I found the source of that chart. And I have a strong suspicion that Eaton is blowing smoke up our asses with this microcontroller story. All that's needed for this to work, is a couple of resistors and a capacitor on the secondary side of the CT sense coil. A high-frequency cut shelving filter circuit.

Anyway, the rationale is that the whole point of GFCI current limitations is to limit the current capable of reaching your heart, and the skin effect attenuates current at depths with increasing amounts as the frequency rises, so the GFCI should be cognizant of that behavior.
 
M

mm08822

Well-known member
Joined
Jan 13, 2012
Messages
5,763
Location
NJ
rlitman said:
Ok, I found the source of that chart. And I have a strong suspicion that Eaton is blowing smoke up our asses with this microcontroller story. All that's needed for this to work, is a couple of resistors and a capacitor on the secondary side of the CT sense coil. A high-frequency cut shelving filter circuit.

Anyway, the rationale is that the whole point of GFCI current limitations is to limit the current capable of reaching your heart, and the skin effect attenuates current at depths with increasing amounts as the frequency rises, so the GFCI should be cognizant of that behavior.
Click to expand...
I provided the Eaton Whitepaper back in post 7. It is not Eaton's data. It is from a study done back in the 40/50's (or further leveraging those insights). Read the 3rd & 4th attachment in post 7.
The data in that chart is contained within UL 943 SB2, so Eaton is just parroting a source.

A low-pass passive element filter alone is insufficient to provide the physical protection to the body as contact with leakage may occur at multiple/other higher frequencies. My understanding from reading is each concurrent frequency component appears to be additive but each at some scalar value that the chart indicates - the "FF" value.

The chart shows 30 frequencies. If we were to break that up into 30 frequency bands, the magnitude of leakage from each band needs to be measured and then the respective FF applied to each. Add those 30 values up and determine if the 60 hz equivalent leakage current is exceeded. If so, trip. If not, no nuisance trip. This is why a microprocessor is needed, otherwise size and drift of components would be an issue. This is how the GFCI becomes cognizant of human safety levels above 60Hz w/o it becoming another PIA device (as some are today).
 
D

dscheidt

Well-known member
Joined
Apr 26, 2017
Messages
2,875
rlitman said:
Ok, I found the source of that chart. And I have a strong suspicion that Eaton is blowing smoke up our asses with this microcontroller story. All that's needed for this to work, is a couple of resistors and a capacitor on the secondary side of the CT sense coil. A high-frequency cut shelving filter circuit.

Anyway, the rationale is that the whole point of GFCI current limitations is to limit the current capable of reaching your heart, and the skin effect attenuates current at depths with increasing amounts as the frequency rises, so the GFCI should be cognizant of that behavior.
Click to expand...
As @mm08822 pointed out, you need some integration of multiple frequencies. But it is worth noting that modern gfci devices already have a microcontroller, as it’s the easiest way to meet the self test requirements that went into effect a decade or so ago. So for a lot of devices, not much needs to change, other than the software it runs. Some devices probably have a blob chip custom to them, that’s a little more work to redo.
 
U

u3b3rg33k

Well-known member
Joined
Dec 18, 2017
Messages
4,037
wyliesdiesels said:
they do not. a GFCI can be installed without connection to an EGC

the pathway for leakage current has nothing to do with what i said.
Click to expand...
to your statement "umm GFCIs dont care about the EGC. they work without them"
I replied "yes they do [work without them]"
 
U

u3b3rg33k

Well-known member
Joined
Dec 18, 2017
Messages
4,037
Chuckster in NJ said:
A GFCI is like a "balance beam" scale…….. What it sees (amps) on the hot wire (3.289 amps) it must see on the neutral wire (3.289 amps) otherwise it will trip. Take a wiggy voltage tester and go from hot to ground and it will trip because there will be a imbalance.
A knob and tube wiring system has no ground and if you replace a receptacle on a knob and tube system it should be a GFI receptacle to be safe.…….. AGAIN, the GFCI does NOT rely on a ground wire for proper operation.
Click to expand...
i'm pointing out that if I remove the EGC from a VFD the GFCI won't trip. even though the VFD is DESIGNED to put the "noise" on the EGC via it's EMC filter, not to mention that the PWM nature of a VFD means there's common-mode noise that won't zero sum by virtue of the design.

one could say a VFD is the perfect tool to nuisance trip a "GFCI" even though there's no ground-fault present.
maybe if I used a big class A amplifier as a VFD I could run it on a standard "GFCI".
 
M

mm08822

Well-known member
Joined
Jan 13, 2012
Messages
5,763
Location
NJ
Higgins said:
Glad to see we are ahead of the curve!
Now How do we, general public know when we purchase a new appliance that a "newer" GFI is required!
Click to expand...
You either change it out as part of the planned installation or play the wait and see game.

Maybe appliance install instructions will get updated with a confirm the installed/new gfci is HF rated.
 
OP
C

Chuckster in NJ

Well-known member
Joined
Jan 26, 2010
Messages
2,265
Location
Hunterdon County NJ
Higgins said:
Glad to see we are ahead of the curve!
Now How do we, general public know when we purchase a new appliance that a "newer" GFI is required!
Click to expand...
Just plug it in and see if you get any nuisance tripping……. "IF" it trips a few times just upgrade the GFI to a new HF type and if it does trip just let it go.

BTW! Hopefully the "general public" has enough sense to call a licensed electrical contractor instead of trying to a "DIY fix" or remove the GFI protection.
 
M

mm08822

Well-known member
Joined
Jan 13, 2012
Messages
5,763
Location
NJ
W

wyliesdiesels

Well-known member
Joined
Aug 14, 2012
Messages
19,968
Location
Modesto, CA
D

dscheidt

Well-known member
Joined
Apr 26, 2017
Messages
2,875
U

u3b3rg33k

Well-known member
Joined
Dec 18, 2017
Messages
4,037
wyliesdiesels said:
go for it. youll get fried... hope you have a defibrillator nearby

30ma is for equipment protection not life safety/personnel
Click to expand...
and yet i've been working with legacy 3ø 240 delta for years and have yet to die. not a single GFCI to be seen.
let's bring the hyperbole down to like a 6?
Honestly I wouldn't be surprised if the higher current rated ones STILL trip, since the 4mA ones trip out even under 4mA which is why I ask.
 
W

wyliesdiesels

Well-known member
Joined
Aug 14, 2012
Messages
19,968
Location
Modesto, CA
u3b3rg33k said:
and yet i've been working with legacy 3ø 240 delta for years and have yet to die. not a single GFCI to be seen.
let's bring the hyperbole down to like a 6?
Honestly I wouldn't be surprised if the higher current rated ones STILL trip, since the 4mA ones trip out even under 4mA which is why I ask.
Click to expand...

so you shock yourself on a regular basis and dont die? that makes no sense
and 240D isnt any more spicy than split phase 120/240 unless youre messing with the stinger leg. anyone who grabs the stinger leg and touches neutral or ground at the same time has whats coming to them....

still dont get your point
 
U

u3b3rg33k

Well-known member
Joined
Dec 18, 2017
Messages
4,037
wyliesdiesels said:
so you shock yourself on a regular basis and dont die? that makes no sense
and 240D isnt any more spicy than split phase 120/240 unless youre messing with the stinger leg. anyone who grabs the stinger leg and touches neutral or ground at the same time has whats coming to them....

still dont get your point
Click to expand...
i see the hyperbole is still at 11.

I just wanna know if the higher current units still insta-trip when the output stage stars. I suspect they would.

also 240 delta is definitely more spicy than split phase. it's 240V to ground. there's no stinger leg. it's 240-240-240.
edit: it's grounded B.
 
W

wyliesdiesels

Well-known member
Joined
Aug 14, 2012
Messages
19,968
Location
Modesto, CA
u3b3rg33k said:
also 240 delta is definitely more spicy than split phase. it's 240V to ground. there's no stinger leg. it's 240-240-240.
edit: it's grounded B.
Click to expand...

aww ok you didnt specify

regardless are you purposely touching an ungrounded phase and the grounded phase? i dont get it :dunno:
 
You must log in or register to reply here.
To avoid these ads, REGISTER NOW!
Top Bottom