GFCI vs RCD trip threshold-why?

[Ingenieur]

So tripping within 6 Hz must mean tripping on any voltage system between 54Hz and 66Hz?

6Hz equals six cycles *per second*, which is not a unit of time. The GFCI is characterized by either the time or the number of cycles till trip.

The transition from cycles to Hz was complicated by the fact that people carelessly called our normal power system 60 cycle instead of 60 cycles per second.

To move this to a more familiar context, would you be OK with a maximum stopping *distance* specified in feet per second?

Sent from my XT1585 using Tapatalk

60 Hz basis
6 Hz/60 Hz = 0.1 sec

 

[wwhitney]

60 Hz basis
6 Hz/60 Hz = 0.1 sec

Still a units error: 6 Hz / 60 Hz = 0.1 (dimensionless).

If you want the result to be a time, you need to say "it will trip in 6 cycles". As in 6 cycles / 60 Hz = 0.1 seconds.

Cheers, Wayne

 

[Ingenieur]

Still a units error: 6 Hz / 60 Hz = 0.1 (dimensionless).

If you want the result to be a time, you need to say "it will trip in 6 cycles". As in 6 cycles / 60 Hz = 0.1 seconds.

Cheers, Wayne

6 out of 60 on 60 Hz basis
0.1 sec

 

[wwhitney]

6 out of 60 on 60 Hz basis
0.1 sec

Still no good. If you want to combine something with "cycles/second" to get an answer in "seconds", it needs to be in "cycles" (or 1/cycles). So the terminology you want is "trips within 6 cycles (at 60 Hz)".

If this still isn't clear, may I suggest brushing up on dimensional analysis?

Cheers, Wayne

 

[Ingenieur]

Still no good. If you want to combine something with "cycles/second" to get an answer in "seconds", it needs to be in "cycles" (or 1/cycles). So the terminology you want is "trips within 6 cycles (at 60 Hz)".

If this still isn't clear, may I suggest brushing up on dimensional analysis?

Cheers, Wayne

60 Hz basis
6 Hz or 1/10th is 0.1 sec

suggest to your hearts delight

 

[wwhitney]

60 Hz basis
6 Hz or 1/10th is 0.1 sec

Nope, sorry, 6 Hz out of 60 Hz is, as you way, 1/10th. Dimensionless.

You can't just add the 'seconds' unit back for no good reason.

If you disagree, please explain why. Posting the same wrong statement repeatedly does not advance the discussion.

Cheers, Wayne

 

[Ingenieur]

Nope, sorry, 6 Hz out of 60 Hz is, as you way, 1/10th. Dimensionless.

You can't just add the 'seconds' unit back for no good reason.

If you disagree, please explain why. Posting the same wrong statement repeatedly does not advance the discussion.

Cheers, Wayne

60 Hz basis
1/10 of the 1 sec basis = 0.1 sec
I need not explain
your repetative stuff does not advance the gfci discussion which is the actual topic of the thread
In the context utilized I'm pretty sure most interpreted 6 Hz as 1/10th sec
if not you clarified it :lol:

 

[mbrooke]

I actually think this is the crux of the matter for us here in North America. The difference between death and injury is the stuff of wet dreams for lawyers. Think "Mcdonalds hot coffee"... They make FAR FAR more money on injuries than they do if someone dies. Sounds sick, but it's true, based on the cost of the victim living with the injury for a long time rather than a dead person who incurs no more expenses. A friend of mine was in a car accident where she lost the front half of her foot, she got $4 million. The 60 year old driver of the car next to her was killed, his family got $600k.

So I can see that if GFCIs were allowed to go to 30mA, the mumber of deaths may not increase, but the number of lawsuits would skyrocket. Even if only a muscle lock took place, the concept of emotional pain and suffering would be very costly for insurance companies. Remember, UL = Underwiters Laboratories, as in INSURANCE underwriters.

The real solution is of course tort reform, but as long as we keep electing lawyers to public office, that is not going to change. Elsewhere in the world, lawsuits like that are not as easily won, in some places you cannot even sue to begin with. 30mA survives as a standard because it doesn't cost anyone much.

Thats an interesting perspective

 

[mbrooke]

UL only makes back reference to NEC for class-A only, not sure why UL even mentions NEC. class-C can be used where some class-A is currently used, but NEC has chosen to not allow class-C within its code.

Makes sense.

 

[wwhitney]

your repetative stuff does not advance the gfci discussion which is the actual topic of the thread

Agreed, so let's drop it. A speedier resolution would have been for you to correct your units error.

Cheers, Wayne

 

[mbrooke]

Where can I find the trip curves for North American GFCIs?

 

[FionaZuppa]

Earth has nothing to do with the function of the test button.
Earth has nothing to do with how the difference in current is sensed by the device.
In the US, 'ground' and 'reference plane' are common synonyms.

Have you asked UL to clarify their language?

under the NEC what else is there? a floating ground is nothing more than bonding. "grounded" is literally "Ground. The earth" per NEC art 100.

 

[FionaZuppa]

Where can I find the trip curves for North American GFCIs?

i will assume this means T_max
at 6mA the max time to trip is 5.593sec
T in sec
I in mA

this seems "normalized" at 20mA, @20mA must trip less than 1sec
@30mA must trip less than 0.560sec


we all know a GFCI class-A test button will trip way way faster 5.5sec, but this is the curve.

 

[mbrooke]

i will assume this means T_max
at 6mA the max time to trip is 5.593sec
T in sec
I in mA

this seems "normalized" at 20mA, @20mA must trip less than 1sec
@30mA must trip less than 0.560sec


we all know a GFCI class-A test button will trip way way faster 5.5sec, but this is the curve.

If true the max for a 6ma GFCI is the same as the typical curve for a 30ma RCD? :?

 

[FionaZuppa]

If true the max for a 6ma GFCI is the same as the typical curve for a 30ma RCD? :?

UL 30mA is about 560ms, the below AS/NZS std is a tad more aggressive.
NHP says:

The trip sensitivity (Rated Residual Operating Current) of the RCD is the minimum level of earth

leakage current required to operate the device. The selection of the trip sensitivity needs to take
into account the purpose of the RCD.
10mA RCDs may be used in areas of increased risk such as circuits supplying outdoor equipment,
bathrooms, kindergartens and hospitals. However, the standing leakage current from appliances
may cause nuisance tripping.
30mA trip sensitivity is the minimum requirement for providing personnel protection whereas
RCDs with lower trip sensitivity (ie. 100mA or above), may be selected to avoid unwanted nuisance
tripping when installing the RCD for the purpose of equipment protection or protection against
fire hazards due to persistent earth fault currents not tripping the overcurrent protective device.

Q:
What is the tripping time for general type 30mA RCDs as per standards AS/NZS
61009 & AS/NZS61008?
A:
As per the table below:
Test leakage current
Tripping times
IΔn 30 mA ≤300 ms

IΔn 60 mA ≤150 ms

IΔn 150 mA ≤40 ms

I don't want to hijack a GFCI thread already in progress but I still want to ask. Why are GFCIs (RCDs) outside of North America set to trip at 30ma while GFCIs in the US are set to trip at 6ma? Isn't 30ma dangerous being past the let go threshold?

why do you say set to 30ma, they make 10ma versions.

 

[mbrooke]

UL 30mA is about 560ms, the below AS/NZS std is a tad more aggressive.
NHP says:

Awesome! But don't most Us GFCI trip much faster?

why do you say set to 30ma, they make 10ma versions.

30ma is the most common and the minimum required by code most of the time. They certainly do make 10ma, 50ma, 100 ma, 500ma and a host of others for various reasons (usually as a main to a TT supply) but 30ma is what you get in most homes and socket circuits.

 

[FionaZuppa]

Awesome! But don't most Us GFCI trip much faster?

they sure do, but all we have for assurance of the product is the UL stamp.

i do not see the makers advertising how fast their crud trips as a way to earn your biz.

much of this issue of differences could come to an end if this stuff was defined under an ISO

 

[mbrooke]

they sure do, but all we have for assurance of the product is the UL stamp.

i do not see the makers advertising how fast their crud trips as a way to earn your biz.

much of this issue of differences could come to an end if this stuff was defined under an ISO

But I thought UL already defined a trip curve for US GFCIs?

 

[FionaZuppa]

But I thought UL already defined a trip curve for US GFCIs?

NEMA & UL are north America standards, an ISO is global std.
the UL trip curve is not a device tested trip curve, its a performance curve to test against, etc. the device must perform at or better than the UL trip curve to be worthy of the GFCI UL stamp, etc.

 

[FionaZuppa]

edit time was up

UL / NEMA are NA std's. an ISO would be better.

my HVAC unit had a failure, a possible fire hazard, UL said they did not certifiy the unit, Canadian CSA did. i submitted to CSA the issue and got nothing back. this issue could be affecting thousands of York made HVAC units.

basically, improper sizing of a relay. one relay pin basically vaporized off during the inrush amps at turn-on. i have since modified the ckt board, it now uses two hefty solid state relays to control the motor. the small relay is rated 10A, inrush amps at turn-on is around 30A !! the large relay was about max'd out, both were swapped to mosfet SSR's protected from flyback by big MOV's. even some of the wiring was bad, had to replace wiring inside the unit as well as the inadequate 14awg cap cord that was feeding the unit, its now hard wired using 12awg into a fused blade disco.