Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Lyle Williams]

150A = 7.5x 20A breaker current for an instantaneous trip.

Thousands of false trips are worth a life saved, but after half a dozen false trips the GFCI/RCD might simply no longer be used.

If mandatory, 6mA seems safer than 30mA.  If not mandatory, a 100mA trip is probably safer than a 1mA trip.

[Tim McCulloch]

A surprising number of folks including too many in our industry have no idea what/how GFCI works, only that they trip when there is no *obvious* visible reason.  Since the user doesn't sense or feel a current leakage (the unit trips to prevent it), they presume the unit is defective and find a way to bypass its use.  They don't realize that lives are being saved and actively defeat the protection in order to use their defective cords, appliances and equipment.

[John Roberts {JR}]

150A = 7.5x 20A breaker current for an instantaneous trip.

apples and oranges... many faults protected by GFCI will not blow the circuit breakers.

Thousands of false trips are worth a life saved, but after half a dozen false trips the GFCI/RCD might simply no longer be used.

That is why I suggested carrying a spare GFCI power drop... If one outlet misbehaves, power up the suspect equipment through the back-up GFCI. If it trips also, you are extremely unlikely to have two faulty GFCI at the same time, so the gear is most likely faulty. (They sell combo extension cords with GFCI ends.)

If mandatory, 6mA seems safer than 30mA.  If not mandatory, a 100mA trip is probably safer than a 1mA trip.

Not sure I follow... 1mA threshold would probably result in more ghost trips, so perhaps more avoidance. 100mA would not prevent meat puppets from getting stuck and injured worse, if not killed.

I don't completely understand the cause of all ghost trips... I saw one occur coincident with a nearby lightning hit, so perhaps a voltage spike or distorted waveform on the mains (WAG?).

If there is enough of a market for seeing less ghost trips these could be engineered to ignore brief faults, but I suspect this response time is specified in the regulations. If anything design effort seems to be targeting self-test and never ignoring a valid fault.

JR

[Stephen Swaffer]

One mA would probably too low a threshold to even be usable-however I hate to think that I am only going to be as safe as someone else makes me be.

There are a lot of safety rules and guidleines that I don't understand-and even a lot of things in code that I think are overkill-but after working in industrial maintenance and as an electrician for a couple of decades there are fewer things that I think are overkill than I used to.

I know sometimes politics plays into it, but most of the thresholds and rules in the codebook are based on statistics-many of them gathered within the insurance industry-but statistics boil down to experience-experience gathered from a much wider playing field than my little corner of the world.

[Guy Holt]

Thus far this discussion has focused on trip thresholds.  It is important to understand that UL943 stipulates a current-time relationship.  The duration that an individual is exposed to fault current is as important as the amount of fault current. To protect against harmful shocks UL943 requires a more rapid response as fault current increases or persists. For example, if the fault current is greater than 300 mA an almost instantaneous response time (no more than 20 ms) is required. If the fault current is only 6 mA a trip delay of up to 5.59 seconds is permitted. The advantage of such an inverse trip curve is that it minimizes nuisance tripping from transient low-current faults while providing protection from ground-fault current.

If there is enough of a market for seeing less ghost trips these could be engineered to ignore brief faults, but I suspect this response time is specified in the regulations. If anything design effort seems to be targeting self-test and never ignoring a valid fault.

There are Class A GFCIs, like the Littelfuse Shock Blocks or Bender Life Guards, that are engineered to ignore fault currents of a sufficiently short duration or high enough frequency so as not to pose a hazard.  The problem is that the hardware store dongle type GFCI most of you are using are not.  Even though the UL 943 inverse-time curve was meant to enable GFCIs to operate more reliably in real world conditions, manufacturers of lower-priced hardware store dongle type Class A devices do not implement the curve because it requires sophisticated micro-processors, which makes the design more complicated and more expensive. Instead they use a more aggressive response (like that illustrated in the figure below) that is lower and faster than that required by UL 943 (typically 250 ms at 6 mA where UL 943 permits 5.59 seconds.)

The more aggressive response of these GFCIs is permissible because the UL standard is the absolute highest current vs. time response accepted, but it is not mandatory. While this more aggressive trip curve does not generally pose a problem in one-tool per circuit applications for which they are meant, it has proven to be a problem in the more extensive distribution of multiple loads that characterizes motion picture production and event staging.

I don't completely understand the cause of all ghost trips...

Ghost trips are usually caused by high frequency residual currents that many electronic devices  shunt to ground on top of minor ground-faults in a distribution.  While individually each of these do not pose a hazard, combined they can easily approach, and possibly even exceed the 5 – 6 mA trip threshold of Class A GFCI devices. Or, use up so much of the GFCI’s allowed leakage current that it becomes excessively sensitive to momentary transient leakage currents caused by the switching on and off of electronic devices. Hardware store type of GFCI are particularly prone to tripping under these circumstances, because of their quick response time and the fact that they do not incorporate high frequency filtration to eliminate nuisance tripping from high frequency residual currents.

To make matters worse, as a consequence of recent code revisions, generator manufacturers  are putting hardware store type GFCIs on portable generators smaller than 15kW.  These GFCIs do not use the more lenient inverse-time trip curve permitted by UL 943, thus making them prone to ghost tripping.

This makes mitigating the effects of residual currents and small ground leaks an important part of ground-fault protection, so that, should an individual take a shock, there is an operational GFCI in place that will cut power within the necessary microseconds. For tips on how to accomplish this with small portable generators see my Protocol series on ground fault protection that is available at http://www.screenlightandgrip.com/html/hd_plug-n-play_pkg.html.

Guy Holt, Gaffer
ScreenLight & Grip

[John Roberts {JR}]

Thus far this discussion has focused on trip thresholds.  It is important to understand that UL943 stipulates a current-time relationship.  The duration that an individual is exposed to fault current is as important as the amount of fault current. To protect against harmful shocks UL943 requires a more rapid response as fault current increases or persists. For example, if the fault current is greater than 300 mA an almost instantaneous response time (no more than 20 ms) is required. If the fault current is only 6 mA a trip delay of up to 5.59 seconds is permitted. The advantage of such an inverse trip curve is that it minimizes nuisance tripping from transient low-current faults while providing protection from ground-fault current.

There are Class A GFCIs, like the Littelfuse Shock Blocks or Bender Life Guards, that are engineered to ignore fault currents of a sufficiently short duration or high enough frequency so as not to pose a hazard.  The problem is that the hardware store dongle type GFCI most of you are using are not.  Even though the UL 943 inverse-time curve was meant to enable GFCIs to operate more reliably in real world conditions, manufacturers of lower-priced hardware store dongle type Class A devices do not implement the curve because it requires sophisticated micro-processors, which makes the design more complicated and more expensive. Instead they use a more aggressive response (like that illustrated in the figure below) that is lower and faster than that required by UL 943 (typically 250 ms at 6 mA where UL 943 permits 5.59 seconds.)

The shock block and life guard solutions look nice but a little expensive for weekend warriors or small sound companies. I noticed one was a rental system, any ball park cost for the big boy GFCI?

The engineer in me says a compromise could be reached that cost far less than those precision engineered, specialty products, while delivering a bunch of the benefit (less false trips). But this would need to jump through the same UL testing hoops, so perhaps too much of a development cost hurdle for what is still probably a niche product.

These days GFCI include self test and turn themselves off when they reach end of life? So more ways to shut down. I guess GFCI outlets are the new razor blades that get consumed by facilities. At least one of my spurious trips was from a self-testing outlet...

The more aggressive response of these GFCIs is permissible because the UL standard is the absolute highest current vs. time response accepted, but it is not mandatory. While this more aggressive trip curve does not generally pose a problem in one-tool per circuit applications for which they are meant, it has proven to be a problem in the more extensive distribution of multiple loads that characterizes motion picture production and event staging.

Ghost trips are usually caused by high frequency residual currents that many electronic devices  shunt to ground on top of minor ground-faults in a distribution.  While individually each of these do not pose a hazard, combined they can easily approach, and possibly even exceed the 5 – 6 mA trip threshold of Class A GFCI devices. Or, use up so much of the GFCI’s allowed leakage current that it becomes excessively sensitive to momentary transient leakage currents caused by the switching on and off of electronic devices. Hardware store type of GFCI are particularly prone to tripping under these circumstances, because of their quick response time and the fact that they do not incorporate high frequency filtration to eliminate nuisance tripping from high frequency residual currents.

Thanks but no light bulb went off illuminating the vector for my several ghost trips... It doesn't happen often enough for me to worry about it. For live sound reinforcement perhaps more of a concern. I still like the idea of carrying back-up to work around and/or confirm a squirrely GFCI outlet/drop.

JR

To make matters worse, as a consequence of recent code revisions, generator manufacturers  are putting hardware store type GFCIs on portable generators smaller than 15kW.  These GFCIs do not use the more lenient inverse-time trip curve permitted by UL 943, thus making them prone to ghost tripping.

This makes mitigating the effects of residual currents and small ground leaks an important part of ground-fault protection, so that, should an individual take a shock, there is an operational GFCI in place that will cut power within the necessary microseconds. For tips on how to accomplish this with small portable generators see my Protocol series on ground fault protection that is available at http://www.screenlightandgrip.com/html/hd_plug-n-play_pkg.html.

Guy Holt, Gaffer
ScreenLight & Grip

[Don Gspann]

As has been discussed on here before, since equipment leakage currents add, the most trouble free would be to have a dedicated GFCI for each 120 VAC circuit.  You could even go to each device, but keep in mind that the GFCI only protects cables and equipment after it, so an extension cord plugged into a non-GFCI receptacle with a GFCI adapter at the end would be a hazard.

OK, so how about these? Kind of a GFCI at the end of an extension cord also?..

http://www.legrand.us/passandseymour/gfcis-outlets/portable/box-portable-gfci/15a/ps154s.aspx

[John Roberts {JR}]

OK, so how about these? Kind of a GFCI at the end of an extension cord also?..

http://www.legrand.us/passandseymour/gfcis-outlets/portable/box-portable-gfci/15a/ps154s.aspx

Yup, there are multiple manufacturers making versions of these.. (check out a Lowes or Home Depot).

And maybe carry a spare...

JR

[Stephen Swaffer]

"Hardware store" GFCIs are no doubt built with the thought in mind that there will usually be one or maybe two appliance plugged in to them so the low threshold is rarely a problem.

I would be curious about GFCI breakers?  There is little info (other than complies with UL943) usually given-probably for liabilities sake.  GFCI breakers presumably would likely be used with multiple appliances.  They are quite a bit more expensive than receptacles as well-but it might make sense to use them in a distro-IF the trip threshold is high enough.

[Lyle Williams]

It has been hard to find numbers, but it looks like annual workplace electrical fatalities per 100,000 workers are:

UK 0.4
AUS 1.0
USA 1.4

There will clearly be many factors at play in these numbers.

[ProSoundWeb Community]