>
>>
>>Early GFI circuits used to use TWO toroid coils with the current carrying
>>conductors run thru the cores.  One coil had a high frequency oscillator
>>driving it, the other was the pickup coil.  An inbalance in the current
>>carrying conductors coupled energy from one toroid to the other.  As far as
>>I know, that system worked down to DC.
>>
>>From what I remember the GFI circuits utilize AC only.

Yep - you are correct.  I looked in my old nat semi data books last night -
they show 2 toroids with each having both conductors run thru the core: one
toroid is at 1000 turns and used to sense differential current, the other
toroid is at 200 turns and is used as part of an oscillator circuit to
detect a grounded neutral.  Since the differential current sensor is by
nature AC coupled, it won't work at DC.  My appologies for my inaccuracy in
the original message.

My best guess is to use something suggested by someone else on this list:
use a large hi-mu toroid core with a gap and install an analog output hall
effect sensor in the gap.  Wind as many turns of your load wires as will fit
on the core.  Subtract the offset on the hall efect sensor with no
differential current flowing.

You could also try a compass module as a high sensitivity magnetic detector.
You would have to mount it in a small heavy-wall steel box to isolate it
from the earch's magnetic feild.

dwayne


Dwayne Reid   <dwayner@planet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(403) 489-3199 voice     (403) 487-6397 fax