PI discrimination

[green]

I've been playing with PI coil on time trying to discriminate ferrous/nonferrous. Start charge, delay 8 counts, integrate preamp out 8 counts(A), delay 16 counts, integrate inverted out 32 counts, stop charge. Compare polarity change of integrator out(C) and preamp out during the 8 counts(A). Opposite for nonferrous, same for ferrous. Seems to work for all the targets I tried except 1 layer of aluminum foil. The first sample would have to be sooner. The 2 outputs were connected to a pot with the wiper to the integrator for ground balance. Seemed to work ok for the clay and the lava rock I have. I have about 40 small ferrite beads in a bag which I tried. Moving the bag caused a polarity change. Not a lot of change at the integrator. Integrator out was measured with a volt meter when plotting targets. A lot would depend on ground balance and probably some things I'm missing

[Tepco]

All this observed thru balanced coil system, then coil will “discriminate”. Try with OO coil, you will get definite polarity, independent of object orientation, this “phasing” is characteristic for DD coils. Logical outcome: square wave driven PI, logical problem: how to ground balance it. Most probably, first sample as soon as possible, last one at the end of cycle, before polarity change, and one in the middle for ground balance.

[Carl-NC]

It's easy to see and distinguish the various target responses during the turn-on time, at least on the bench. The trick is to distinguish them in the presence of a variable ground signal.