Print

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

[Ryan C. Davis]

Had a power issue at a show last night.

Their power distribution is a home made cabinet fed by a 50A GFI breaker (over some #6 SO Cable)  which then gets distributed to 6 separate GFI outlets fed by their own 20A breaker. Power measured right at 120V. These shows are usually small enough that I can run all of my gear off of a sing 20A circuit fed into my Furman PM-PRO. I usually run my mixer off of an extension cord connected the Furman just to try and avoid any ground loops, etc... Never had a problem doing it this way.

Last night I started with my regular setup and every time I would fire up the mixer (Presonus 24.4.2) it would pop the 50A main breaker. Totally puzzled. I tried a couple of things including unplugging the amps and DSP to see if that made a difference an the 50A breaker would still pop whenever the mixer was turned on. I finally moved the mixer to it's own outlet in the distro panel and had no issues after that. The mixer has such a light draw that I can't imagine it was an amperage issue but I thought maybe having the distro outlets being GFI's fed by a GFI breaker maybe that could cause some issues? Hell I don't know what was going on but it was certainly odd.

Anyone seen an oddity like this?   

[Mark Cadwallader]

If the mixer was the only thing plugged into the Furman, then the Furman would be the prime suspect if everything worked once you took the Furman out of the configuration.

[Stephen Swaffer]

I wouldn't be cascading GFCIs-there is nothing "magic" going on inside a GFCI-and no way that I know of to know if there is a GFCI either downstream or upstream of a GFCI-short of tripping a GFCI. 

I've not done a lot of troubleshooting on 2 pole GFCIs-I am surprised that the 50 A breaker was tripping first, but that could just be a little faster trip threshhold on that specific breaker that is within production specs.  My guess is that the leakage currents of all of the devices added up to just enough to trip the GFCI.  Moving the mixer to a seperate receptacle probably put it on the other phase of the 50 A breaker.
With it on the other phase, the leakage currents would have subtracted from each other taking it below the trip threshold.   

[Ryan C. Davis]

If the mixer was the only thing plugged into the Furman, then the Furman would be the prime suspect if everything worked once you took the Furman out of the configuration.

The Furman was never taken out of the system, it was still powering 2 amps and DSP. The problem only occurred when the mixer was plugged into the furman. I even tried unplugging all of the amplifiers and the DSP and even in that scenario with only the mixer plugged in it still tripped the main breaker.

[Mike Sokol]

The Furman was never taken out of the system, it was still powering 2 amps and DSP. The problem only occurred when the mixer was plugged into the furman. I even tried unplugging all of the amplifiers and the DSP and even in that scenario with only the mixer plugged in it still tripped the main breaker.

Sometimes gear with the lowest current draw can have the highest line-to-ground leakage current. For instance, a cheap MOV protected "surge" strip will often leak 2 or 3 mA to ground. You might get away with one of those on a GFCI, but string two of them together and you're right on the edge of tripping the GFCI. EVERYTHING leaks a little, and it's additive.

Let's put our heads together and come up with a cheap/simple way to read line-to-ground leakage currents on individual pieces of gear. I've got a $1,000 Fluke meter for that will go below 1mA, but cheap ($50) clamp-on ammeters usually bottom out at 10 mA resolution. I have a few ideas that should work but I'll need your help to try them. Anyone else have experience in sub-mA clamp ammeters? Maybe this should be a simple inline ammeter? 

[Geoff Doane]

With it on the other phase, the leakage currents would have subtracted from each other taking it below the trip threshold.   

It's only a guess, but I think Stephen has nailed it with this answer.  The 50A GFCI is likely designed for hot tub use, where it's very important to have a low trip threshold.  For our uses, since the 50A feed can't be accessed directly by the public (no regular outlets), it probably doesn't need to have a GFCI at all, or if it does, the threshold could be set higher since there is lower level protection farther downstream.

GTD

[John Roberts {JR}]

I am repeating myself by now, a separate gfci power drop/extension cord could be used to troubleshoot suspicious gear/outlets fed from a different power source.

If two GFCI with the same leakage threshold are in series, the +/-1 mA tolerance could decide which one trips first.

JR

[Mike Sokol]

I am repeating myself by now, a separate gfci power drop/extension cord could be used to troubleshoot suspicious gear/outlets fed from a different power source.
JR

Yes, you are repeating yourself.... Yes you are repeating yourself....

That's why you really should run a separate GFCI power drop for each musician on stage, if possible. That way the bass player's aging rig with the leaky power transformer won't trip the backline GFCI outlet and take out power to the guitar player's rig. They hate that.

And here's me repeating myself. Ground-fault leakage currents are additive. Ground-fault leakage currents are additive. That's why it's a bad idea to have a single 50-amp hot tub GFCI protecting your entire stage. Far better to have half a dozen power drops, with each 20-amp drop protected by a GFCI. Then when something trips a GFCI, it only takes down one instrument on stage. Much easier to troubleshoot during/after the show. and much less impact during the show itself. It's a win/win situation (is that me repeating myself?)

[Frank Koenig]

Let's put our heads together and come up with a cheap/simple way to read line-to-ground leakage currents on individual pieces of gear.

I built myself a little ground leakage probe by opening up a hardware store GFCI receptacle and bringing the secondary leads of its differential transformer out to some banana jacks. The rest of the GFCI circuitry is abandoned. An AC voltmeter (or scope) plugged into the banana jacks measures leakage. Calibration is performed using a known resistor from leg to ground.

I don't know about its frequency response, which would be useful for investigating SMPS induced nuisance trips, but for simple 60 Hz leakage it works fine.

--Frank

[John Roberts {JR}]

It is pretty simple to unplug the suspect gear and measure resistance from one pin of the plug to chassis ground with a VOM.  If unit is turned on the two plug pins will be low impedance to each other.

This may very slightly understate AC leakage, but will probably be good enough to identify a problem unit.

In manufacturing we perform the leakage test at elevated voltage.

JR 

[Stephen Swaffer]

YThat's why it's a bad idea to have a single 50-amp hot tub GFCI protecting your entire stage. Far better to have half a dozen power drops, with each 20-amp drop protected by a GFCI. Then when something trips a GFCI, it only takes down one instrument on stage. Much easier to troubleshoot during/after the show. and much less impact during the show itself. It's a win/win situation (is that me repeating myself?)

I agree-however the 2017 NEC requires all outdoor 50 amp single phase receptacles to be GFCI protected.  Indoors you will OK-but outdoors it's gong to be a problem sooner or later.

[Frank Koenig]

It is pretty simple to unplug the suspect gear and measure resistance from one pin of the plug to chassis ground with a VOM.

JR,

True this. The Ohmmeter is the first test. (That's what I suggested for sorting the tingling tubs.) But if the device passes that and still trips the GFCI we need to see what the GFCI is seeing.

I built that fixture when trying to sort out nuisance trips with some Powersoft K-10s. They tripped the GFCIs in a certain venue, but would not trip any of the GFCIs in my home, or anywhere else I have used them. My fixture revealed a good bit of differential HF hash, which, for lack of a better explanation, I believe is responsible for the problem.

Anyway, I just like building little test fixtures

Best,

--Frank

[Mike Sokol]

I built myself a little ground leakage probe by opening up a hardware store GFCI receptacle and bringing the secondary leads of its differential transformer out to some banana jacks. The rest of the GFCI circuitry is abandoned. An AC voltmeter (or scope) plugged into the banana jacks measures leakage. Calibration is performed using a known resistor from leg to ground.

I don't know about its frequency response, which would be useful for investigating SMPS induced nuisance trips, but for simple 60 Hz leakage it works fine.

--Frank

Frank, what's the ratio of leakage current to measured voltage across the test resistor? You knew I was going to ask that, didn't you?

[Frank Koenig]

Frank, what's the ratio of leakage current to measured voltage across the test resistor? You knew I was going to ask that, didn't you?

Hi Mike,

With a line voltage of 122 V and a test resister of 17.5 kOhm I should have a leakage current of 7 mA. I, somewhat arbitrarily, selected a burden resister of 1 kOhm, across which I get 5.4 mV under these conditions. The calibration appears to be pretty linear for a factor of 2 up and down.

The short-circuit current (as close to a short as the AC-uA range of the Fluke meter gets) is  6.1 uA. The DC resistance of the secondary is about 50 Ohms. We're dealing with pretty small signals in any case, and need a good AC meter.

--Frank

[Jonathan Johnson]

Let's put our heads together and come up with a cheap/simple way to read line-to-ground leakage currents on individual pieces of gear. I've got a $1,000 Fluke meter for that will go below 1mA, but cheap ($50) clamp-on ammeters usually bottom out at 10 mA resolution. I have a few ideas that should work but I'll need your help to try them. Anyone else have experience in sub-mA clamp ammeters? Maybe this should be a simple inline ammeter?

Make up a short cord with male and female plugs. In between, loop the ground wire 10 times.

Then use your cheap clamp-on ammeter with the 10 mA resolution to measure the current on the ground wire, and divide by 10 to get the current.

Of course, this assumes that ALL the leakage current returns on the ground wire.

You could also loop both the hot and neutral wires together 10 times and measure that instead (a differential current transformer), but I'm not certain of the safety of that idea in terms of the ability of the loop to dissipate heat. Of course, you'd need to leave the ground out of THAT loop.

However you do it, the wiring and loop of the cord would need to have an outer jacket to protect from physical damage.

[Mike Sokol]

Make up a short cord with male and female plugs. In between, loop the ground wire 10 times.

Then use your cheap clamp-on ammeter with the 10 mA resolution to measure the current on the ground wire, and divide by 10 to get the current.

Of course, this assumes that ALL the leakage current returns on the ground wire.

Yep, I done that and it works. If you only have one piece of gear plugged into test, then the ground current should be the total leakage current. If the gear is connected to other gear via XLR connectors and such, then you could have secondary current paths. But for testing a single piece of suspect gear on the bench that's probably the most cost effective way since you can use a $50 clamp-on ammeter.

[ProSoundWeb Community]