Can the IB-PI differentiate between metals?

[Tinkerer]

Can the IB-PI tell the difference between metals?

I started a new thread for this test series, because the gain and settings are not exactly the same as the other test series.

Here are six coins of very similar size, but made of different alloys.
The pictures show the signal during TX ON to the left and after Switch OFF to the right.
Looking only at the signal after the switch OFF, does not tell much more than amplitude.
Looking at the signal during TX ON, also shows FE, the signal goes positive but also shows differences between non-FE alloys.

First picture: DIFFERENTIATING TEST ALL TARGETS
#1 Copper-Nickel-US$-10c-Dime

Tinkerer

[Tinkerer]

Ooops, forgot to post the NO Target picture. So here it comes.
Differentiating Test No Target

And then #2. This coin looks like Nickel on the outside, but you can see that the signal to the left has gone positive, so there must be a FE core. Indeed, it does stick slightly to a magnet.

Tinkerer

[Tinkerer]

Target # 3 is a copper coin, 1c from Trinidad Tobago. Looks like a straight forward copper coin.

[Tinkerer]

Target # 4, an old Venezuelan 5c coin. It looks like Nickel on the outside, but it sticks to a magnet.
You can also see the FE content on the left signal response.

Tinkerer

[Tinkerer]

The last one of this series.


# 6 is a modern 5c from Venezuela. It looks like copper on the outside, but it sticks to a magnet. You can also note the FE content on the left signal trace.

Tinkerer

[Tinkerer]

Goofed up.

Here is # 5 It is a 5c coin from Switzerland. I don't know what the alloy is, but probably it is copper and nickel.

So that completes the series. You be the judge. Does the IB-PI tell the difference between the coins of similar size?

Tinkerer

[simonbaker]

To answer the question whether sampling during TX-on allows superior discrimination, someone needs to give a really good try at using just the after-TX data to do discrimination also. Are we sure the information isn't in there?

Of course including sampling during TX-on gives you twice as much data, so that is a plus right there.

What is the theory as to why samples during TX-on provide unique information? Is it the different dphi/dt that is important, or something to do with the static magnetic field that persists after dphi/dt goes to zero?

If it is simply the different dphi/dt that extracts more information, could we get a similar capability by alternating different damping resistors in the TX-off pulse to create two different kinds of waveforms to sample and still use a single coil?

Regards,

-SB

[Qiaozhi]

If it is simply the different dphi/dt that extracts more information, could we get a similar capability by alternating different damping resistors in the TX-off pulse to create two different kinds of waveforms to sample and still use a single coil?

When I experimented with a mono versus an induction-balanced coil, probably over one year ago, I discovered that the phase information associated with the target is only detectable with the IB. My conclusion was that the mono is effectively being driven by a forcing function (i.e. the transmitter) and any superimposed target phase info is completely swamped by the TX. But in the IB arrangement the TX signal is suppressed by the coil arrangement, allowing the target phase info to be seen.

[Tinkerer]

When I experimented with a mono versus an induction-balanced coil, probably over one year ago, I discovered that the phase information associated with the target is only detectable with the IB. My conclusion was that the mono is effectively being driven by a forcing function (i.e. the transmitter) and any superimposed target phase info is completely swamped by the TX. But in the IB arrangement the TX signal is suppressed by the coil arrangement, allowing the target phase info to be seen.

I agree. It also let's you see the target's response at the time it happens.
Looking for the response after the Flyback has subsided and the circuit has settled, can only show you a vague shadow of the original signal.

Tinkerer

[simonbaker]

PI power req

What are the power requirements of a PI detector compared to a VLF detector?

-SB

[Aziz]

Hi simonbaker,

What are the power requirements of a PI detector compared to a VLF detector?

-SB

too much worse!

On the VLF detectors, the oscillating resonant tank LC has minimum losses compared to the PI detector. The magnetic field energy will be converted in defined cycle time (frequency) into electric field energy in the capacitor and then vice versa. During this conversion, some losses occur. The oscillator just injects as much energy to the LC tank as the losses were. A very big part of the stored energy will be recycled.

On the PI detectors, all of the magnetic field energy will be converted into heat on every cycle. The stored energy won't be recycled as on the VLF's occur. This will of course increase the global warming a bit.

Aziz

[Tinkerer]

What are the power requirements of a PI detector compared to a VLF detector?

-SB

500mW for TX is not bad for a PI
1 Watt for the TX pulse is quite a powerful PI.
I consider using 2 Watt for the TX pulse for a very powerful IB-PI

For the rest of the circuit I don't think there is too much difference.

Tinkerer

[Tinkerer]

When I experimented with a mono versus an induction-balanced coil, probably over one year ago, I discovered that the phase information associated with the target is only detectable with the IB. My conclusion was that the mono is effectively being driven by a forcing function (i.e. the transmitter) and any superimposed target phase info is completely swamped by the TX. But in the IB arrangement the TX signal is suppressed by the coil arrangement, allowing the target phase info to be seen.

It would be very interesting to hear all the other advantages and disadvantages that you found, when experimenting with the IB coil.
There is so little information on the forums about IB coils.

Tinkerer

[simonbaker]

Hi simonbaker,



too much worse!

On the VLF detectors, the oscillating resonant tank LC has minimum losses compared to the PI detector. The magnetic field energy will be converted in defined cycle time (frequency) into electric field energy in the capacitor and then vice versa. During this conversion, some losses occur. The oscillator just injects as much energy to the LC tank as the losses were. A very big part of the stored energy will be recycled.

On the PI detectors, all of the magnetic field energy will be converted into heat on every cycle. The stored energy won't be recycled as on the VLF's occur. This will of course increase the global warming a bit.

Aziz

I ask because I am wondering if the PI-IB design is really a VLF detector in disguise. If we were to put the same power into a VLF, maybe it would compete more with PI.

What do I mean by PI-IB is VLF in disguise? Well, by sampling during TX turn-on and after TX, it is like looking at one cycle of a continuous TX wave form. A chain of TX pulses has a fourier signature with a low frequency sine wave and harmonics. The fundamental frequency is not too useful, but we are interested in some intermediate harmonics for doing discrimination.

VLF takes two samples by multiplying RX signal by phase offset square-waves of same frequency - estimating phase of RX signal.

By sampling PI pulses during TX and after, you can get data probably similar to phase. It is a little like turning a synchronous detector on and off for little periods during the TX pulses.

But is it a very sensitive way to do it? Individual sampling in PI is probably much noiser than a VLF synchronous detector method. However, PI pulse has so much power, we think it overwhelms noise better than VLF - but what if we put same power into VLF?

-SB

[nick_f]

Hi Simon and Aziz,
I also thought myself that if we can generate a sinewave in a VLF with a peak of a few hundred volts (as the flyback in PI), wouldn't we get the same sensitivity as with the PI? Besides, Aziz perfectly explained why the VLF has a much better power efficiency than PI.
But if the PI-IB is similar with VLF, it will be plagued by the same problems with bad ground and it will require GEB. Is there a way to identify if Tinkerer's detector will be affected by bad ground? How about testing with a piece of ferrite, is that similar to bad ground?
Does anybody have a performant VLF, I would like to see some results from detecting a 1 square inch and half an inch of aluminium foil (and specify the frequency).

Regards,
Nicolae