Tests of big depth metal detectors ( TR and PI systems )

[Davor]

You could add a parallel capacitance in order of 50pF in parallel to Tx and Rx coils to account for parasitics, and the real life cable capacitance. The damping resistors will most certainly get smaller to achieve good damping.

[green]

Looks good.
As you see, you can't simulate the thickness-dependent behaviour that is relevant for PI. I made a viscous soil model, but so far I did not try simulating the thickness.

Correct, don't know how to simulate skin effect. Click run, click C, click amplitude scale(change to log). The coil on decay slope is the same as the coil off decay slope similar to the Coke can recording that I got on the bench. Spice TC=30usec, Coke can TC about 60usec(thin so not much skin effect).

[bbsailor]

You could add a parallel capacitance in order of 50pF in parallel to Tx and Rx coils to account for parasitics, and the real life cable capacitance. The damping resistors will most certainly get smaller to achieve good damping.

Parasitic properties should always be included in any modeling. If you add parasitic TX / RX capacitance, add coax cable capacitance and MOSFET output capacitance you definitely will need a lower value damping resistor to optimize the damping time but you will also change the full stimulation of very small targets due to the coil discharge TC being one fifth target TC rule of thumb.

This is very much like the difference between ideal transformer theory and real transformer theory. It is always good to consider how the real detecting environment affects what you can expect in detecting performance. Here is an example. Detecting on a wet salt water beach makes the wet sand respond below about 15 uS. This makes finding smaller gold targets more hard to find in wet sand but easier to find small gold targets in dry sand at lower delays.

There are classes of targets and target environments that require us to optimize our designs on these real conditions and knowledge of these interactions in: 1. Optimum coil size for a class of target TC ranges and sizes, 2. TX pulse time and power needed to fully stimulate a target, 3. Optimum delay time to detect a particular target, 4. The response of the environment to being stimulated in an attempt to detect a target, 5. Noise in the detecting environment.

Joseph J. Rogowski

[Davor]

Well, yes and no. Real life parasitic effects in many cases overwhelm the root causes of a phenomenon, so in order to learn about it you may only benefit from parasite-free learning environment. Silly as it may seem, adding parasites one by one leads to better engineering, and a better grasp of what areas desperately need improvement.

I think I have a candidate for a thick target. It has octave related time constants, and only one coil is being excited/interrogated. I did not check it against PI excitation. Hope it works.

[green]

Well, yes and no. Real life parasitic effects in many cases overwhelm the root causes of a phenomenon, so in order to learn about it you may only benefit from parasite-free learning environment. Silly as it may seem, adding parasites one by one leads to better engineering, and a better grasp of what areas desperately need improvement.

I think I have a candidate for a thick target. It has octave related time constants, and only one coil is being excited/interrogated. I did not check it against PI excitation. Hope it works.

Tried your target with spice(PI). Maybe similar to a US quarter reply #33.

[Davor]

Looks promising. This model is set with 3 octaves, and my idea was to fit the real life behaviour of a given target by tuning the inductances and resistances. It could lead to a better explanation of the "holes" in PI GB schemes.

[Funfinder]

MDs are using radiation, and higher radiation causes higher and longer distant effects.
So the solution is simple:

Using 30 volts @ 10.000mA together with 70cm coils should be a good start
instead of ridiculous "deep search experiments" with 9v and 30cm coil MDs!

Heavy mineralization only is a problem for weak detectors and for
such "beepers" like the Garrett Ace where someone can't distinguish
between erratic-noise- and real find-beeps.

Many so called detectorists are just an idle and spoilt bunch of
"park-walkers" who want it as soft, easy, nice and relaxed as possible.

No wonder if there is no true development going on if all
these hobbyist-freaks prefer those weak coin-finders!
For the real good MD's they have no understanding at all,
same as partially here, in this mainly Pulse Induction forum!

[Ferric Toes]

MDs are using radiation, and higher radiation causes higher and longer distant effects.
So the solution is simple:

Using 30 volts @ 10.000mA together with 70cm coils should be a good start
instead of ridiculous "deep search experiments" with 9v and 30cm coil MDs!

Heavy mineralization only is a problem for weak detectors and for
such "beepers" like the Garrett Ace where someone can't distinguish
between erratic-noise- and real find-beeps.

Many so called detectorists are just an idle and spoilt bunch of
"park-walkers" who want it as soft, easy, nice and relaxed as possible.

No wonder if there is no true development going on if all
these hobbyist-freaks prefer those weak coin-finders!
For the real good MD's they have no understanding at all,
same as partially here, in this mainly Pulse Induction forum!

None of what you say makes any sense and is very disparaging of detector users and developers. It's like saying everybody should be buying 500HP supercars, even if it is only to go to the local shops, or the school run. You won't get much radiation from 30V and 10mA, or should that dot be a comma?

Eric.

[Phil S]

I wish I could just walk and dig the park as I'm sure it's full of decent sterling silvers

[Koala]

None of what you say makes any sense and is very disparaging of detector users and developers. It's like saying everybody should be buying 500HP supercars, even if it is only to go to the local shops, or the school run. You won't get much radiation from 30V and 10mA, or should that dot be a comma?

Eric.

The more current the bigger the battery.

Lets swing with a car battery strapped to your back.

[Skippy]

Wouldn't the detector detect the battery every sweep, with it not being attached to the machine? Seriously, there's plenty of folks who WOULD use a machine like that if it actually did the business and found stuff.

Back on-topic, I've just finished casting a lead disc, 60mm diam. x 5mm , as detailed in my earlier post. I hope to be able to bodge together my main machine soon, and then I'll try and make some time to do some proper science stuff.

[lytle78]

This forum is unique. Nowhere else is there such a lively discourse between those knowledgeable in the art of detector design, enthusiastic amateurs and total whack jobs.

Hope no one is offended.

But - to all of us - keep it up - it bears strange and occasionally valuable fruits.

[green]

Wouldn't the detector detect the battery every sweep, with it not being attached to the machine? Seriously, there's plenty of folks who WOULD use a machine like that if it actually did the business and found stuff.

Back on-topic, I've just finished casting a lead disc, 60mm diam. x 5mm , as detailed in my earlier post. I hope to be able to bodge together my main machine soon, and then I'll try and make some time to do some proper science stuff.

Interested in your lead disk testing. Been doing some more penny testing. The pennies I have test different. 1962-1982 pennies are 95% copper 5% zinc, 1982 to present are 97.5% zinc 2.5% copper. Sorted the ones I have, mostly 1982 and higher. Did get one 1982 that tested higher and six 1982 that tested lower. Charted TC and some integrator out values. You suggested using pennies and the lead disk as possible targets in one of your replies. Thought the coins made more sense at the time. Now I'm thinking 3 or 4 different lead disks would be better for standard test targets, maybe 5, 15 and 45mm thick depending how your lead disk tests. The lead disks should be more repeatable. There was some comparison testing using 5 and 6mm squares cut from a soda can awhile back. It was hard to cut to desired size. Bigger squares not so hard. If I'm going to make a lead disk the same as yours it would help to have the same form to pour the lead. Maybe a can or something else? The coins were placed random, in a plastic container trying to not have any on edge. A lot of pennies have a higher TC than one but not a lot higher. Thinking 50, 150, 450 and maybe 1350usec TC might be some thing to try for. Just thinking maybe all wrong. Maybe someone already down this road could suggest something.

[Skippy]

I did suggest using copper (pre-1982-ish) US 1 cents for several reasons. One was for consistency, if someone else is to reproduce a test, they don't want to do so with a random mix of zinc-cored and copper coins. Secondly, the copper coins are the close match to our U.K bronze 1 pence coin (we went to a copper-plated-steel coin in the 80's, for the same economic reasons as the US went to zinc-core). Also, I understand 'zinc' cents corrode badly in the ground, this makes them less suitable as a test target, im my opinion.

Re: the lead disc... the size/weight I've started off with (60 x 5, 160gm) was an engineering guess, based on a number of ideas. Clearly it's significantly larger than any coin-size object, so that puts in in the range where it should be detectable deeper than any coin. If you're searching in a way that reduces pickup of small coin-size items (eg. coil clear of the ground), then it will still be visible. During my own real-world deep-searching trials, I did detect a golf-ball sized lump of lead weighing 150 grams... it was only about 10 inches (30cm) down, so would probably have been detected with normal search methods. I had the search-coil about 40cm above the ground when searching.

It may well turn out to be a bit too large and easy to detect, but that's the point of experimentation, you've got to start somewhere.

I like the lead disc as a target because it could be representative of a real target - some coin hoards have been found buried in lead pots, which are often fragmented. Plus it could also be oriented vertically or at an angle to give that 'coin-on-edge' effect (see Riss' first post).

Regarding my mould for the lead: My first attempt was made from thick aluminium foil (actually 0.07mm) as used for meat pies, pastries etc by food retailers. I found some jam/preserve which had a lid about the correct size, and shaped the foil around it, creasing up the edge. It was about 12mm deep and 58mm diameter when I'd finished, close enough. The correct weight of clean-ish lead was weighed out before melting, allowing 5 grams extra for losses, eg. dross, metal staying in the melting pot etc. The mould was placed on a ceramic tile for thermal reasons.
I suppose I should state: take precautions. Gloves, eye protection, well-ventilated area, water-soaked cloth to hand. etc etc.

[Skippy]

Attached a pic of the lead disc and mould.