tightly wound copper coil is bad ?

[Tec]

Iron is a piece of Al-foil hiding in a Magnetic field

Didn't have much luck with the toroid.

I did get a harmonic 'ringing' happening through it, but it doesn't change much, and no noticeable or useful difference between Fe and non-Fe.

Yes scope observations have been very helpful..

I am happy to report that Reliable Fe discrimination is possible with one coil.

[Tepco]

Harmonics and ringing you observed is probably because toroid (current transformer) is not terminated whit proper resistor on its secondary side.

Fe discrimination whit one coil is possible, BFO, PLL, FM,off resonance,even monocoil VLF(?!?) machines can do this, but what about straight PI at maximum range?

(Never mid, if i'm not right, you will be reach!)

PS: I will add some more comments (more like idea check or opinion pool for everyone here) as soon as possible, without making new topic about it, will be OK here, not OT.

[Tec]

Fe discrimination range

Dunno about maximum range..

Can the other technologies mentioned do Fe discrimination at maximum range ?

From my observations, the iron effect is from 5% to 20% of the total signal.

If this is the case, balanced systems, or BFO or other mentioned technologies would not be able to discriminate at the limits of their detection range.

Is this the case ?
... that a VLF detector can not tell the difference between iron and non-iron if the target is deep or faint.

The best I aim for is to match the discrimination abilities of a induction balance (VLF) system.

[Tec]

The magic ratio is the comapraison of the signal at two different sample delays. Early, and later on. Giving a indication of 'conductivity'. Not complicated, or special. .. Just the signal at say 15uSec, compared to at 25uSec. i.e Measure the 'conductivity' of the target at different pulse widths. The bottlecap looks like a coin, because it is shaped like a coin. What you actually see is the shape of the target, different eddy currents becoming more or less prominent at different coil pulse widths. That is all.. so a complicated item like a screw or bobby pin, has a complicated curve, because at certain pulse widths a certain eddy current path will be stronger. The coins and simple shapes have one dominate eddy current path..the copper coin reaches a peak at the pulse width that corresponds nicely with it's eddy current timing. (see the graph). I don't really need to measure a ratio to see this, I could have just measured the signal at a certain sample delay, just like any pi machine and plotted it vs different pulse widths.
Anyway, I am working on an android interface right now. An android phone cost 60 bucks, this will be the display, touch interface, and controller for the PI system. Please let me know if anyone wants to pre-order the system. If you don't have an android phone, I can sell you one (for $60).
Tec