Distinguish aluminum trash from gold jewelry?

[Diga]

Hello. First time poster here.

Is it possible for an induction balance metal detector to distinguish between gold rings and aluminum trash? Is there any way A.I. would be beneficial for such?

Thanks.

[Carl-NC]

Simple answer is No. The benefits of AI is very debatable in metal detectors. In order for it to be useful, it needs to know exactly what target you just dug, how deep, at what angle, and also the loss angle of the ground mineralization. And a pull tab with a flat beaver tail, a missing beaver tail, and a bent-over beaver tail will all have slightly different IDs. The amount of variables are enormous. I won't say it'll never be done, but AI in detectors will certainly be a challenge.

[pito]

https://www.youtube.com/watch?v=BseNR36b8BE&t=18s
my attempt, rejects not only aluminum

[Diga]

Simple answer is No. The benefits of AI is very debatable in metal detectors. In order for it to be useful, it needs to know exactly what target you just dug, how deep, at what angle, and also the loss angle of the ground mineralization. And a pull tab with a flat beaver tail, a missing beaver tail, and a bent-over beaver tail will all have slightly different IDs. The amount of variables are enormous. I won't say it'll never be done, but AI in detectors will certainly be a challenge.

Thank you for the reply Carl.

Other than the amount of ground mineralization, a detector doesn't, and can't, know all those other variables. Even if it magically did know those variables, I don't think it would matter. For example, a pulltab on edge will have a certain signal signature, but there are countless other good and bad targets that would share the same signature. Correct or incorrect?

[Carl-NC]

but there are countless other good and bad targets that would share the same signature.

Basically correct, but by using multiple frequencies and taking data throughout the target sweep (not just the peak) there is more information that might distinguish some similar targets. It's basically a massive real-time data collection exercise.

[ivconic]

... or make any of the classic Tesoro designs, TGSL, Eldorado... Bandido (recently made one) and adjust geb&disc so precise to reject aluminum trash and accept some golden items... it is possible, only Tesoro can do it like that!
Only problem with Tesoro circuit is the masking problem. Above mentioned will really work fine only on clean and medium polluted soils, but will totally fail in heavy mineralized and polluted soils.
It all depends on exact personal preferences and subjective conditions on a given site. Sometimes even a simple Tesoro circuit will do the job and sometimes even ML GPX 6000 will fail completely.
Until Carl, Moodz, Altra, Tinkerer and few others figure out some ultimate solution and materialize it into product for market... looking forward!

[Smirnov-Arta]

...

Good afternoon. Speaking of gold or aluminum products, do not forget that their chemical composition is significantly different. We just say gold, but really? For example: gold products of 958 samples (95.8% pure gold, the remaining 4.2% are a mixture of silver and copper) or gold products of 585 samples (58.5-59% of gold contains another 7.5-8.5% silver and 32.5-34% copper), etc. Similarly with aluminum products. This leads to difficulties in identifying some products from others. In addition to the mineralization of the soil, first of all it is necessary to understand its magnetic viscosity...the number of ferromagnets and paramagnets in the soil. Sorry for the Google translation.

[ivconic]

True points!

[Skippy]

And don't forget, aluminium is not all the same, either.
Pure ( or nearly-pure ) aluminium is what foil ( chewing gum wrapper, cigarette packet ) is made from, and that has an IACS conductivity about 65% of copper ( = 100% ). But drinks can metal is a very fancy high-strength alloy, and has IACS of typically 37%. This affects how they appear to a detector.

One thing that may help distinguish the two is a mechanical pinpoint probe. Drinks can bits ( canslaw ) can give a distinctive sound. That may help you avoid digging them .... but then they remain in the ground until next time you visit that location, and detect them again. And repeat.

[ivconic]

The problem is related to a subjective set of conditions.
On "my" hills where there are ancient sites; there are no major problems with Al waste, of any type and grade.
Another problem is mostly present there; the presence of a large percentage of ferrooxide in the scattered material as well as high mineralization of the soil.
Aluminum waste of various types and alloys is present mostly in lowlands, river banks, and banks with arable land.
I get totally discouraged sometimes in places like that.
Because aluminum in various shapes, thicknesses, masses provides a range of audio and VDI response that completely covers the range of desired findings.
Deus is totally "confused" in such places. More, thicker pieces of aluminum give a very strong signal, which is often misleading and I think it's a shallow finding.
And then I "die" from digging a huge hole, at the bottom of which I find some kind of garbage.
Considering that with every plowing, the soil "overturns" and every year some new discoveries "emerge"; in such places it is highly desirable that the detector is "shallow" and not "deep".
But in addition to that, it also has strong and precise discrimination.
And now I come to the story of Tesoro detectors.
Ideal in such conditions! They have no competition!
Because they are extremely "shallow" and at the same time, with precise adjustment, GEB and DISC can very successfully separate aluminum debris from ancient coins.
Of course, Deus can be set to behave similarly. I have a special customized program "Fast2", which is a modified version of the stock program "Fast".
Disc is at "10", Sense at "92", GEB from case to case, Recovery speed at "2.5-3" and in my opinion the most important Silencer function is at "4".
But even so set; it often happens that Deus is deceived and I dig up some shallow Alu garbage.
But Tesoro is NEVER deceived if it is set up precisely.
In this case, I first rely on years of experiences with TGSL, Eldorado and Bandido.
There... Tesoro is not "dead", it is especially usable in such conditions.
The northern part of Serbia is called "Vojvodina", a huge plain, the "granary" of Serbia.
Ancient Roman coins, as well as a mass of various silver coins from later periods, are most often found in the fields there.
Tesoro is the most common "tool" among experienced searchers out there.
Here, first of all, I am talking about very experienced seniors who have been in this hobby for over a couple of decades.
I do not base my views, statistics and opinions on the basis of beginners, of which there are more and more.
Because they haven't even scratched the surface of the essence of this hobby yet.
I'm a senior and I know a lot of seniors. And all that I claim; first of all, I base it on such experiences.
As I write this I suddenly have a "flashback" to the wonderful TV series "The Detectorists"...
There is a lot of truth in that series!