[LAU] ground loop hell

David Christensen dpchrist at holgerdanske.com
Mon Feb 9 02:46:38 UTC 2015

On 02/08/2015 05:35 PM, Bob van der Poel wrote:
> ​Just wondering on all this ... where the grounding is "suspicious" does
> the use of a ext. chord with a built in GFI switch add any safety? Probably
> doesn't solve grounding loop problems.

Wikipedia has an article on residual-current devices (RCD):


Whether using an RCD is more or less safe depends upon what you're 
doing.  If you're using a power tool with bad insulation while standing 
on wet concrete, a RCD could save your life.  But if the RCD is powering 
something critical (ventilator, chemical reactor, etc.) and the RCD 
trips, bad things could happen.

The only interaction between RCD's and ground loops I can think of under 
normal conditions (e.g. no ground fault) would be whether or not the RCD 
passes or omits the equipment grounding conductor.  I seem to recall 
seeing early residential GFI receptacles with only two prongs (hot, 
neutral).  Everything I've seen in the past ~20 years has all three 
(hot, neutral, ground).  The former would break ground loops and the 
latter would allow them.

Something to understand is that RCD's can nuisance trip -- e.g. trip 
when you don't want them to.  For example, there is one GFCI receptacle 
in my kitchen that protects all the receptacles.  It trips whenever I 
plug or unplug a particular counter-top appliance; sometimes 
immediately, sometimes later.  If I forget to reset the GFCI, my wife's 
phone doesn't charge and I hear about it.

Here is an audio example -- I was doing live stage sound an at outdoor 
event with many acts each day.  One 120/240V "spider" box with GFCI 
circuit breakers and receptacles powered everything -- sound, lights, 
and whatever the acts connected.  A band took the stage, started 
performing, and the power started cycling in and out.  I had to stop the 
show and trouble-shoot it.  The problem turned out to be the keyboard 
player's surge suppressor outlet strip (and possibly my equipment 
rack-mount outlet strip).  Those strips contain three non-linear 
semiconductor devices connected across hot, neutral, and ground.  At low 
voltages, say ~200 VAC, they have high impedance.  But at high voltages 
(e.g. a surge), they conduct:


The problem is that there is always leakage current, and the spider box 
GFCI's saw that as a ground fault.  It was curious that the power cycled 
in and out, rather than simply tripping off.  The solution was to remove 
the keyboard player's surge suppressor outlet strip and use a plain 
outlet strip.


More information about the Linux-audio-user mailing list