Print

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

[Guy Holt]

I suspect that since many switching power supplies include line-to-chassis RF caps to reduce their own high frequency trash, that it's possible for those same capacitors to act as a shunt for an external transient... Is that what's working its way back into these GFCI sensing circuits and causing them to trip "randomly"?

That is exactly what I found in tests that I did for a series of articles on ground fault protection strategies with portable generators that I wrote for Protocol magazine. Like so much modern electronic equipment today, the electronic ballasts of many movie lights eliminate RFI by incorporating a mains input RF filter.  The RF filter snubs the high frequency harmonics created by the ballast to the EGC. Even though the capacitor values of the filter are chosen so that they conduct at high frequencies, a tiny amount of current at the mains frequency (either 60 or 50 Hz) still flows through the capacitors resulting in leakage of both the fundamental and harmonic frequencies to the equipment-grounding conductor, creating residual currents between 3-4mA.

Since these currents do not return to their source via the neutral conductor they cause GFCIs to see a difference between the current leaving on the hot line and the current returning on the neutral line. If this residual current does not trip the GFCI outright, it can reach the point where transient surges in power will trip the GFCI. This typically occurs when the sum of residual leakage reaches approximately 30% of the GFCI's rated sensitivity threshold (1.5mA.) Once GFCIs have become "sensitized" by residual leakage current, transient events can result in surges of residual current that cause them to trip. For examples of transient events that may push a GFCI over the edge and cause it to nuisance trip see my Protocol articles (links available at http://screenlightandgrip.com/html/hd_plug-n-play_pkg.html.)

Guy Holt, Gaffer
ScreenLight & Grip
[email protected]

[Michael Storey]

I've had similar issues just about every time I've had to run my iTech 4k/8k's hard off of a GFI outlet. I believe there's just something about their design that causes this.

[Henry Cohen]

. . . Like so much modern electronic equipment today, the electronic ballasts of many movie lights eliminate significantly reduce RFI by incorporating a mains input RF filter . . .

Fixed that for ya.

[John Escallier]

With every gust of wind through the building the series LED indicators on the Shock Blocks would run up into the red zone indicating an increase in leakage current and then run down as the wind subsided.

Guy Holt, Gaffer
ScreenLight & Grip
[email protected]

Interesting observation, thanks.

What bugs me in the conclusion however, is this:  When the wind dies down, the humidity is still there.  As such, the loss of wind cannot be responsible for the loss of leakage by the effect of humidity.

Since you see that high humidity is needed as well as wind, I suspect the effect is a combination of entities.

Humidity increases the conductivity of otherwise insulating surfaces.  That in itself is sufficient for the riggers to get the shocks.  I suspect that the shock breaker you used is not indicating actual leakage, as that would mean that the guys would still get shocks.  Instead, I suspect (this is a WAG on my part) that the led indicator is displaying voltage potential.  Wind will cause triboelectric potentials, and I suspect the meter is high enough impedance to see that DC potential as well as the AC leakage.

I would have to review the schematics of the unit to confirm this, unless someone here has a good understanding of the circuitry.

I have to deal with humidity and leakage constantly.  Every time we hipot midsummer, we have the problem.  I end up bagging components at both ends (50 foot long 3 foot diameter vacuum cryostats), and blowing dry nitrogen through for a day or two.  Trying to keep leakage at 5kV down to 100 nA.

John

[Jonathan Johnson]

Not by building up a charge.... I think this gear is being "randomly" turned on at the top of the 60 Hz sine wave and drawing a current pulse with a fast enough rise time to capacitively couple back though the chassis to the EGC wire. And if you plug in a piece of gear with its power switch already on, you'll create a lot of arcs and sparks with high-frequency content. I think that's whats causing the GFCI's to trip....

Maybe some of the other posts have addressed this in language that I don't quite understand, but here goes.

Capacitors are used to correct for power factor -- when current and voltage do not align. So, when powering up some devices, is it possible that some capacitance (inductance?) or power factor issue is causing the current on the neutral to lag or lead the current on the hot? That would mean that while the RMS input/output current is the same, the waveform of the current is not aligned between the hot and the neutral. The GFCI senses this misalignment as an imbalance and trips.

As I understand, GFCIs must trip within 1/40th of a second (25 ms). But in order to prevent tripping due to a power factor issue, it seems to me that a slower trip might be needed: perhaps as long as two cycles (or 1/30th of a second - 33 ms).

I'm not an engineer; I'm looking forward to the EEs' explanation of this.

[Guy Holt]

What bugs me in the conclusion however, is this:  When the wind dies down, the humidity is still there.  As such, the loss of wind cannot be responsible for the loss of leakage by the effect of humidity.

Spoken like a land lover. Quite often, in the evening, a wet breeze comes in off the ocean.  Where the ship yard  was right on Quincy Bay, a damp wind would blow through the building causing the humidity to spike momentarily.

Guy Holt, Gaffer
ScreenLight & GRip

[John Roberts {JR}]

Maybe some of the other posts have addressed this in language that I don't quite understand, but here goes.

Capacitors are used to correct for power factor -- when current and voltage do not align. So, when powering up some devices, is it possible that some capacitance (inductance?) or power factor issue is causing the current on the neutral to lag or lead the current on the hot? That would mean that while the RMS input/output current is the same, the waveform of the current is not aligned between the hot and the neutral. The GFCI senses this misalignment as an imbalance and trips.

Where do you get this?

I am not sure load angle affects GFCI as long as the currents net out over the waveform period (16 msec).

BTW PF is only corrected with caps for modest power level appliances, and perhaps big dog mains power apps. PFC in audio products (like amps) are done with electronic circuitry.   

As I understand, GFCIs must trip within 1/40th of a second (25 ms). But in order to prevent tripping due to a power factor issue, it seems to me that a slower trip might be needed: perhaps as long as two cycles (or 1/30th of a second - 33 ms).

I'm not an engineer; I'm looking forward to the EEs' explanation of this.

Sorry I can't, I'm not an EE... but I did sleep in a Holiday inn once .   

JR

[John Escallier]

Spoken like a land lover. Quite often, in the evening, a wet breeze comes in off the ocean.  Where the ship yard  was right on Quincy Bay, a damp wind would blow through the building causing the humidity to spike momentarily.

Guy Holt, Gaffer
ScreenLight & GRip

Sigh..."lover"  not lubber?

Been livin within a mile of the Atlantic pond from the day I entered this world.. Navy brat from day one, Chealsea, newpawt, Lon-guyland... Hole in the wall, big green monster, Faneuil Hall... No beaner gonna talk to me that way...

Your anecdotal account seems to be too fast.  The air is not the conduction path, from zero RH out to 100%, it's the surfaces.

I can see at some temperature, the breeze coming in above or below the dew point, but cannot see adsorption based modulation of surface conduction varying up and down that fast.  The closer to dewpoint, the slower the system will change.

Absorption slower still.

Given that, I'm still not sure what your meters were reading.  Hence the query.

Me, I woulda fixed the building grounds. Even small buildings like those, you have to worry about the bonding integrity.  Especially when they get old in a seashore environment.

John

[Stephen Swaffer]

Guy,

I appreciate your input-and honestly would like to read your findings regarding transients.  But I would prefer to read about them without digging through many many pages of marketing material that is irrelevant to the topic at hand.  Any chance you "cut and paste" for us?

Are your Shock Stop GFCIs UL listed?  "Hardware store" GFCIs (actually the ones I install come from one of the 2 top names in devices with years of experience building quality devices) are supposed to trip at 6 mA for safety purposes.  A higher trip threshhold would seems to lessen the safety value of the device?

[Guy Holt]

As I understand, GFCIs must trip within 1/40th of a second (25 ms).

That is only because manufacturers cut corners to save money. UL943 (the standard for Class A GFCIs) permits a much more lenient inverse time trip curve than what is used in most GFCIs out there. While researching an article on ground fault protection strategies for Protocol magazine, I came across some tests conducted by Littelfuse (manufacturer of film Shock Blocks) that demonstrate that inexpensive portable GFCIs do not take full advantage of the inverse time trip curve permitted by UL943.

The inverse time relation of the UL943 trip curve is designed to permit transient conditions that are sufficiently short in duration so as not to pose a hazard while keeping current through the body at safe levels. To assure the latter, UL943 requires that as fault current increases the maximum allowable time to open a circuit and interrupt power decreases, with an almost instantaneous response time required (no more than 20ms) if the fault current is greater than 300 mA compared to 5.59 seconds at 6 mA.

Relationship of typical GFCI trip curve to the UL 943 Curve (taken from “Now that industrial GFCIs are here...” by Nehad El-Sherif, Jan/Feb 2014 issue of the IAEI Magazine.)

Even though the UL 943 inverse-time curve is meant to enable GFCIs to operate more reliably in real world conditions, manufacturers of lower-priced Class A devices do not implement the curve because it requires sophisticated micro-processors, which makes the design more complicated and the GFCI more expensive. Instead they use a more aggressive response (like that illustrated above) that is lower and faster than that required by UL943 (typically 25ms at 6mA where UL943 permits 5.59 seconds at 6mA.) 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. That is, a device will fail UL testing if it responds slower than the standard requires, but will pass as long as the response time is less than the curve time even if it is a lot less. 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.

The manufacturers of inexpensive GFCIs also don’t take advantage of the high frequency filtration permitted by UL943 and so their GFCIs become sensitized by residual currents, which makes them even more prone to nuisance tripping.

Left to Right: 100A/120V Shock Block, LifeGuard, & Shock Stop Film Style GFCIs

One solution to noisy loads is to use a film-style GFCI, like the LifeGuard, Shock Block, and Shock Stop GFCIs pictured above.  To deal with the harmonics that Kino & HMI ballasts draw, and achieve something approximating the desired Frequency vs. Trip Current Curve of UL943, these manufacturers incorporate harmonic filters that attenuate harmonics. Attenuated by the filter, the effective trip level of the harmonic currents generated by dirty loads is substantively increased thereby reducing the incidence of residual currents causing a GFCI to nuisance trip.

50A/240V Shock Stop Film Style GFCI

This combination of high frequency filtration and a more lenient trip curve enables film-style GFCIs to provide more nuisance free ground-fault protection. For more information on the use of film style GFCIs see my Protocol articles available at

Part 1: http://www.screenlightandgrip.com/images/Protocol_Art_1_.pdf
Part 2: http://www.screenlightandgrip.com/images/Protocol_Art_2_.pdf
Part 3: http://www.screenlightandgrip.com/images/Protocol_Art_3_.pdf

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

[ProSoundWeb Community]