[Prev][Next][Index][Thread]

Re: Ground Fault Interupters and RFI filters





---------- Forwarded message ----------
Date: Mon, 29 Sep 1997 15:37:12 -0400
From: Thomas McGahee <tom_mcgahee-at-sigmais-dot-com>
To: Tesla List <tesla-at-pupman-dot-com>
Subject: Re: Ground Fault Interupters and RFI filters 

> ---------- Forwarded message ----------
> Date: Sun, 28 Sep 1997 22:10:34 -0500 (CDT)
> From: Richard Wayne Wall <rwall-at-ix-dot-netcom-dot-com>
> To: tesla-at-poodle.pupman-dot-com
> Subject: Ground Fault Interupters and RFI filters 
> 
> 9/29/97
> 
> I recently added a 30 A circuit breaker/GFI to my power cabinet. 
It's 
> placed before the main contactor and variac.  Following the variac 
> there is a 100 A RFI filter.  Without the main power transformer 
> connected, the GFI trips when power is applied.  It still trips
when 
> either the neutral or hot is disconnected and the other remains 
> connected.  Removing the RFI filter stops the GFI tripping.  I
replaced 
> the RFI filter with two 200 A EMI filters on both the neutral and
hot 
> line.  Same thing happens as with the RFI filter.
> 
> Has anyone had this experience with GFIs and RFI filters?  Anyone
know 
> what the mechanism is?
> 
> RWW

Richard,

Since there is no load, when you apply power by throwing the switch,
the capacitors on the hot side of the RFI filter will experience a
change in voltage that will cause a current flow through them to
safety ground. This current will NOT flow through the neutral wire,
so the GFI will trip. :(

I assume that you are switching both the hot AND the neutral wire.
That would explain why the circuit trips when there is only a filter
set on the neutral side. We generally assume that neutral and safety
ground are at the same (ground) potential. Tain't necessarily so,
Joe! Let's assume that the neutral side filter has a capacitor
between neutral and safety ground, and that there is NO filter on the
hot side. Further assume that the switch is open and the capacitors
on the neutral side are discharged. Let us further assume that you
have some other appliances plugged in to the same circuit. 

The ground currents in the neutral wire will cause it to to have a
voltage with respect to the safety ground, which normally carries no
current at all. When you throw the switch, this difference in voltage
between neutral and safety ground will cause the RFI filter
capacitors to pass a current. This current will NOT be the same as
the current in the hot circuit. So, the imbalance in current will
trip the GFI. If the neutral line is NOT shared by some other
circuit, then this explanation would not be applicable. But I just
bet that you do have something else connected on that line. Note that
you can eliminate this problem by simply not switching the neutral
wire.

OK, so then why doesn't this same mechanism cause the circuit to trip
the GFI when there IS a load? Because when there IS a load, the
aforementioned transient current spike is SMALL compared to the hot
and neutral currents. The GFI is more sensitive to the transient
currents caused by the RFI filter capacitors charging/discharging
when the TOTAL current flow is large. The mechanism for this loss of
sensitivity when there is a load is due to the way most GFI circuits
sense the current in the first place.

Many GFI devices utilize a toroidal core with three windings. Two of
these connect to the main power lines and carry the full current.
They are wound in such a way that the resulting magnetic fields
cancel one another if the currents are equal. If they are NOT equal,
then the resulting magnetic field couples to the third winding, which
is usually many more turns of fine wire. This produces a voltage that
is proportional to the current difference in the load lines. It is
this signal that is used to trip the circuit. When there are large
currents flowing through the two main windings, then the sensitivity
of the third winding to current imbalance is reduced. The circuit
sensitivity reduces with total current flow. If the current flow
causes core saturation, then sensitivity goes right out the window
and operation can become erratic.

If you are really bothered by the accidental tripping of the GFI with
no load and the RFI filter(s) in place, then just add something like
a 15 watt light bulb as a built-in, always there resistive load. Or
use anything else as a load, if you don't like lightbulbs flooding
the dark with superfluous photons.

Hope this helps.
Fr. Tom McGahee