Print

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

[Lyle Williams]

I think the only place that I've ever seen a fence bonded & grounded is around an electrical substation.

Transmitter sites too.

But bonding and grounding isn't the best we can do.  Tripping a breaker is clearing the fault only once it has reached a clearly lethal level.  We hope that nobody is part of the circuit from when the fault begins until the fault clears.  In some circumstances that time window might be a fraction of a second.  In other circumstances the time window may be years.

GFCI/RCD reduces this time window to milliseconds for some common types of faults.  Regulations should allow GFCI devices with trip currents high enough to minimise nuisance trips.  Lower trip currents might be safer, but not if they result in GFCIs not being installed in the first place.
 

[Mike Sokol]

GFCI/RCD reduces this time window to milliseconds for some common types of faults.  Regulations should allow GFCI devices with trip currents high enough to minimise nuisance trips.  Lower trip currents might be safer, but not if they result in GFCIs not being installed in the first place.

I would agree. GFCIs are routinely bypassed for backline stage power because of nuisance tripping, but that's because code requires us to use the lowest threshold 6mA trip version. I routinely see musicians with the ground pins of their guitar amp broken off to stop the GFCI in the wall from tripping. Of course, that may work when they're plugged into a GFCI outlet, but the next time they're plugged into a conventional outlet they have NO shock/electrocution safety systems in place. So in that case, the nuisance tripping of the GFCI causes them to eliminate their EGC safety grounding system. Yikes...

If "industrial" 30 mA GFCIs were allowed for these applications, then there would still be substantial safety from electrocution, but far fewer "nuisance" trips. And perhaps the safety grounds on the stage amps would stay intact.

The same sort of situation happens in the RV industry where a camper will sometimes use a TT-30 to NEMA 5-20 adapter to plug their RV shore power into a campground pedestal. That's because there's a 6 mA GFCI required on the 20-amp NEMA 5-20 outlet, but not on 30-amp TT-30 outlet. So again, a safety system that could save a life has been bypassed to prevent a nuisance trip that could shut down the refrigerator or water tank heaters while they're away for the day. That random trip can result in the loss of frozen food or cause damage to the RV plumbing system. So again, the GFCI is bypassed because of it tripping randomly. They're seen as a "nuisance", not the piece of life saving thechnology they really are.

I am talking to the code committee at the RVIA (Recreational Vehicle Industry Association) about modifying their build code to allow the use of 30 mA GFCIs for shore power feeding campers, but it's a slow road. Remember, the camper itself should have 6 mA GFCIs in place for outside and kitchen/bathroom outlets already. I'm not arguing that those should be changed to 30 mA threshold. I believe this 30 mA threshold GFCI should only be allowed on the pedestal circuit feeding the distributed power system of the RV which feels the combined leakage of all internal RV systems.

[Guy Holt]

GFCIs are routinely bypassed for backline stage power because of nuisance tripping, but that's because code requires us to use the lowest threshold 6mA trip version…. If "industrial" 30 mA GFCIs were allowed for these applications, then there would still be substantial safety from electrocution, but far fewer "nuisance" trips. And perhaps the safety grounds on the stage amps would stay intact.

In motion picture set power distribution we encountered very much the same problem.  Part of the problem is that GFCIs are quite often used too far upstream where they monitor more than one load. The solution we came up with is to use PFC equipment when ever possible and deploy GFCIs of varying trip levels in an “Interlocking Ground Fault System.” We use PFC equipment because a common cause for nuisance tripping in GFCIs is that manufacturers of electronic devices will quite often capacitively couple high frequency harmonic currents to ground to reduce the amount of RF signals emitted into the atmosphere. Harmonic currents capacitively coupled to ground do not return to the GFCI via the Neutral and so cause the GFCI to see a difference between the current leaving on the "Hot" line and the current returning on the Neutral line. Because the trip level and trip time of Class A GFCIs is very low, such high frequency currents traveling on the equipment grounding wire will trip GFCIs.

Another cause of nuisance tripping is caused by the mutual inductance of conductors arranged closely parallel to one another in what is called "Proximity Effect." At fundamental frequencies, Proximity Effects are usually negligible, but can increase significantly with current frequency. The reason for this is because system capacitance has an impedance defined by 1/(2 x pi x f x C). What this means in practical terms is that the high frequency harmonic currents generated by dirty loads reduces the impedance of the system, which results in an increase of current leakage due to Proximity Effect. As a result, leakage of high frequency harmonic currents (in the 20 Khz range) may be 300 times larger than 60 Hz values.

The graph above illustrates the Proximity Effect  at varying harmonic numbers  (and an analogous problem "Skin Effect") for 12 AWG cable.  From this graph it is immediately apparent that Proximity Effect can be significant in small cables. Within the range of the Triplen Harmonics alone (i.e. 3rd, 9th, 15th),  Proximity Effect increases by 60% in 12 AWG Cable.  In a large distribution system operating a number of dirty loads, the difference between Hot and Neutral currents as a result of high frequency current leakage due to induction, when combined with the Capacitive Coupling of equipment, can exceed the allowed 6mA threshold of Class A GFCIs and cause them to trip. This tripping is considered a nuisance because the current not returning via the Neutral is seen as a problem by the GFCI when in reality both the coupled and induced currents return to its source in a safe and controlled fashion via the equipment grounding wire. Even though this “system noise”  does not generally pose a shock hazard, as Mike has pointed out the nuisance tripping it causes leads users to  circumvent or not use GFCIs.   For this reason, in motion picture set power distribution we use PFC equipment whenever possible in order to minimize the magnitude of harmonics our loads draw.

We also use an “Interlocking Ground Fault System” consisting of GFCIs of varying trip levels because even after eliminating sources of harmonic currents, a large distribution system is still likely to have ground leaks. Capacitive reactance in the generator, a poorly insulated or defective extension cord, defective insulation in a lamp housing, or defective plug, all will leak a little current (to name just a few causes.) While these faults may be small, over a sizable distribution system they can all run together, like many ground springs running together to form a stream, as they all return to the same source - the generator. Under such circumstances, someone touching the generator will receive a shock while someone else touching a lamp with a fault up stream will just experience a tingle.

An “Interlocking Ground Fault System” like that illustrated above, consisting of GFCIs of varying trip levels, can be very effective in providing the Class A personnel protection required by code, while at the same time eliminating nuisance tripping. Employing an Interlocking Zone Ground Fault Protection scheme is very much like employing a zone defense in basketball. The distribution system is broken up into zones wherever there is a change in wire size or branch circuit. A GFCI sized for the over current device used to protect the branch circuit is positioned downstream of the over current device. The result is a cascade of interlocking protective zones starting at the power source and ending at the load. As you can see in the table below, there exist a wide variety of GFCI devices for this purpose - ranging from Class A devices with fixed 6mA trip levels and devices with user adjustable trip settings from 5- to 50mA. Adjustable devices set for a trip level of 20mA meets the UL943 standard for Class C protection of equipment but is not suitable for personnel protection (Note: devices with adjustable trip levels are capable of providing personnel protection to UL943 Class A specifications when set for a trip level of less than 6mA and time delay of 100 milliseconds, but are not Class A devices because the UL Standard for Class A requires a fixed threshold of 6mA).

The forward zone 120V and 250V single phase GFCIs must employ Class A devices with trip settings of less than 6 mA to assure the safety of personnel. As long as that is the case, the rear zone 208/240V multi-phase GFCI can employ higher trip settings (typically 10 or 20mA.) The choice whether to use a trip setting of 6-, 10-, or 20mA depends on the circumstances and the level of protection desired - but some type of rear zone GFCI protection is a good idea for the reason cited above.

Left: Bender 400A 3 Phase Adjustable GFCI. Middle: Shock Block 400A 3 Phase Adjustable GFCI. Right: 100A 3 Phase Adjustable GFCI.

The Bender Corporation offers a device that can be very useful in these situations. The trip level of their RCMA420 ground fault monitor can be adjusted steplessly from 10 mA to 500 mA. An adjustable time delay is also available, from 0 to 10 seconds. And a digital display shows the measured fault current in real-time.  These features enable the user to eliminate nuisance tripping from system noise by enabling them to first assess the level of system noise and then set a trip threshold above the noise floor that will trip in the event of a real fault event. Because the UL Standard for Class A requires a fixed threshold of 6mA, the RCMA420 ground fault monitor can not serve as the exclusive ground fault protection in a system, but if set to a tolerable shock level (less than 20mA with a very short time delay) it can offer a degree of  protection against accumulated ground leakage and eliminate nuisance tripping,  while Class A devices at the individual loads offer the Class A personnel protection required by code. 

For more detailed information on how to design a ground fault protection system for temporary installations, see my IA Ground Fault Protection workshop available at http://www.screenlightandgrip.com/html/481_GFCI_Workshop.html.

Guy Holt, Gaffer
ScreenLight & Grip
www.screenlightandgrip.com

[Frank DeWitt]

Guy and Mike make very good points.  If safety equipment such as GFCIs cause to much trouble people will disable them in order to get the job done.  Should they do it or is it justified is immaterial.  They will.   A neighbor bought a new riding lawn mower and found that he could not mow the slopes on his property because he could not lean to either side as he mowed without the seat safety switch killing the mower.  He returned the mower and had his old one rebuilt.

BTW A Friend of mine discovered that he could not use the PRO on his new farm tractor to run a pump because the tractor would shut off when no one was on the seat He was revealed to find that a second safety device solved the problem.  If he adjusted the seat belt so it put pressure on the seat when fastened directly across the seat then the tractor would run. 

[Stephen Swaffer]

But bonding and grounding isn't the best we can do.  Tripping a breaker is clearing the fault only once it has reached a clearly lethal level.  We hope that nobody is part of the circuit from when the fault begins until the fault clears.  In some circumstances that time window might be a fraction of a second.  In other circumstances the time window may be years.

I agree that GFCIs are helpful-but ultimately they are "engineering solutions" and subject to failure as well.

Bonding puts a low resistance path in parallel with a persons body-even a 6 piece of #12 has a low enough resistance that following ohms law, it would limit a 1500 amp fault to allowing 6 ma to flow if a person's body has 1500 ohms resistance.  Larger wires drop resistance considerably-I would probably never use something as small as #12 myself.

For temporary installs, it would usually be impossible for you to install GFCIs-especially as mains or as Guy suggests (I know Guy's are usually temporary-but I am also considering the typical budget of many guys on here) -but bonding is a reasonable option at whatever level you are at. If you are dealing with stuff large enough that bonding may not give a fast enough response, then an interlocking system such as Guy illustrates is probably within your budget.

[John Roberts {JR}]

Guy and Mike make very good points.  If safety equipment such as GFCIs cause to much trouble people will disable them in order to get the job done.  Should they do it or is it justified is immaterial.  They will.   A neighbor bought a new riding lawn mower and found that he could not mow the slopes on his property because he could not lean to either side as he mowed without the seat safety switch killing the mower.  He returned the mower and had his old one rebuilt.

BTW A Friend of mine discovered that he could not use the PRO on his new farm tractor to run a pump because the tractor would shut off when no one was on the seat He was revealed to find that a second safety device solved the problem.  If he adjusted the seat belt so it put pressure on the seat when fastened directly across the seat then the tractor would run.

One of my early co-op jobs was working QC in a factory that had heavy presses, punches,  and cut-off saws and paid workers by the piece, so they were motivated to work fast. I also saw a number of workers around the factory who were missing parts of digits because of old pre safety rules accidents, or cheating the newer safe machines. One trick was to put your elbow on one of the two safety switches so you could press both buttons with one arm still free...(DO NOT DO THIS).

I have trouble with the safety switches on my riding mower, because dirt or whatever across the switches cause just enough conduction on the kill winding of the magneto to make it hard to start. I added a switch in series with the wiring harness so I can defeat the safety switches just to get it started the first time. After it is started and warmed up I can start it and run with the leaky safety switches active.

JR

PS: I am not enthusiastic about using higher trip current GFCI while it seems appropriate for a whole distro barrier protection. I prefer the idea of outlet strips with GFCI for each small group of outlets, so leakage problems can be easily segregated and identified. I wouldn't want to use a guitar amp leaking more than 6 mA.

[Lyle Williams]

In Australia (where we don't have "safer" 115v domestic supply, just 230v) ordinary GFCI/RCD devices are 30ma, and industrial GFCI/RCD are 300ma.

Is the 6mA trip current in the US too low?  Isn't 6mA the fatal threshold for a current delivered to the heart?  Isn't the heart always wired up in a voltage/current divider we call the body?

If we don't budge on 6mA but use of GFCI/RCD is optional, aren't we actually choosing a 150,000mA trip current?

[Mike Sokol]

PS: I am not enthusiastic about using higher trip current GFCI while it seems appropriate for a whole distro barrier protection. I prefer the idea of outlet strips with GFCI for each small group of outlets, so leakage problems can be easily segregated and identified. I wouldn't want to use a guitar amp leaking more than 6 mA.

I would agree. The real problem is that leakage currents are additive, and even non-obvious devices such as "surge strips" can leak a few mA to the EGC ground wire and still be UL and Code compliant. So one guitar amp on a surge strip plugged into a GFCI on a stringer might not cause nuisance tripping, but two surge strips and two guitar amps on the same GFCI just might get over the 6 mA threshold. What does this mean for those of us doing live sound gigs? Well, I suggest a one-amplifier/one-GFCI policy. So there will be no additive currents causing a nuisance trip, and if one does occur it will only take down one stage amplifier.

But that does suggest that further up the power distro stream (at the generator or sub-panel for stringers) that there's really no GFCI protection if a distro wire gets tangled up in a fence and the insulation gets cut. So perhaps that's the place for 30 mA or 50 mA GFCIs that will protect humans from upstream ground fault incidents such as an electrified fence.

Finally, it would be great to have a simple test for how much leakage current is actually occurring. The problem is that clamp-on ammeters that go down to 1 mA resolution are pretty expensive. I've got one on my bench that costs around $1,000, but that's not something for casual testing. The standard 10 mA resolution clamp ammeter that costs $100 can't tell you if an amplifier is leaking 1 to 5 mA current to ground and getting ready to trip your GFCI. To remedy this I've suggested that Amprobe should make a 10:1 current multiplier specifically for ground leakage testing that could be used with any 10 mA resolution clamp ammeter to give it 1 mA resolution. I've asked their engineering department about this a few months ago, but haven't heard back. So I'll ping them again and see if they're interested in building such a gadget. I've posted this here before, but a basic "line splitter" wired with the ground instead of the hot line in the loop would be useful for GFCI ground leakage testing.

What do you guys think? Would you buy one of these modified to do ground current testing for $20 or so and use it to test back-line amplifiers for ground fault leakage? 

[Stephen Swaffer]

One of my early co-op jobs was working QC in a factory that had heavy presses, punches,  and cut-off saws and paid workers by the piece, so they were motivated to work fast. I also saw a number of workers around the factory who were missing parts of digits because of old pre safety rules accidents, or cheating the newer safe machines. One trick was to put your elbow on one of the two safety switches so you could press both buttons with one arm still free...(DO NOT DO THIS).

Hence the invention of "anti tie down" circuits so both switches must be released between parts and then the next generation "concurrent switching" requiring both buttons to be pressed within .5 seconds of each other.  And yet when they figure out how to defeat that and get hurt they still get a nice settlement from "the company."  So much for education teaching them to be safe.

A device to "break out" ground leakage would be a great troubleshooting tool.  Thinking out loud-a 1 ohm resistor would allow you to probe with a voltmeter with 1 mV=1mA, would that affect the leakage enough to make it inaccurate?  I am thinking a DVM accurate at the mV range might be easier to come by than an accurate current clamp meter at that range?

[Frank DeWitt]

A missing ground line and short circuit from a football field light caused the death of 13-year-old Lincoln Middle School student Christian Lorinczy, the investigation shows.

http://www.mlive.com/news/ann-arbor/index.ssf/2014/10/short_circuit_missing_ground_l.html

More questions.  We do know that bonding the fence and ramp would have prevented the problem. 

And this in the comments.
You are extremely paranoid. Why would anyone carry a tester around all the time unless a problem was expected or they plan on working with electrical appliances.

[ProSoundWeb Community]