A question for Carl_NC or anyone that might know.

[Cherry Picker]

May I pry on your knowledge of detector design for a minute?

I'm still stuck on the idea of a detector that can tell gold from aluminum.

Do any current detectors, that you know of, use current drop of an Eddy current to help determine metal composition?

Thanks

 

[Rudy]

It is hard to figure out what you mean by "current drop of an Eddy current". Do you mean the decay of the Eddy current, after the transmitted field rate of change has stopped?

I believe most induction balanced VLF detectors look at the phase difference between the transmit field and the build up of the Eddy current. A high conductance target exhibits a larger phase difference and a low conductance target exhibits a smaller phase difference. This phase difference is a key component in determining the VDI assigned by the detector. Other factors are important such as: the amount of material, its geometric shape, etc.

In a time domain detector (ie. a PI machine), the Eddy current signal from low conductance targets peaks shortly after being hit with the transmit field and decays quite rapidly, whereas a high conductance target's Eddy current peaks much later and lasts a bit longer. That is why PI detectors with longish minimum delays don't make good gold finding machines.

A small digression:
Although as detectorists we keep talking about "conductance", this is only partly correct. In reality, it is the target's "admittance" that is being evaluated. Admittance has two components, one being conductance, the other being reactance.

 

[Cherry Picker]

It is hard to figure out what you mean by "current drop of an Eddy current". Do you mean the decay of the Eddy current, after the transmitted field rate of change has stopped?

I believe most induction balanced VLF detectors look at the phase difference between the transmit field and the build up of the Eddy current. A high conductance target exhibits a larger phase difference and a low conductance target exhibits a smaller phase difference. This phase difference is a key component in determining the VDI assigned by the detector. Other factors are important such as: the amount of material, its geometric shape, etc.

In a time domain detector (ie. a PI machine), the Eddy current signal from low conductance targets peaks shortly after being hit with the transmit field and decays quite rapidly, whereas a high conductance target's Eddy current peaks much later and lasts a bit longer. That is why PI detectors with longish minimum delays don't make good gold finding machines.

A small digression:
Although as detectorists we keep talking about "conductance", this is only partly correct. In reality, it is the target's "admittance" that is being evaluated. Admittance has two components, one being conductance, the other being reactance.

Thanks Rudy,

I know they use the same system with the Eddy current in sewage plants to identify and separate the metals. I have seen several that specifically said they, using the Eddy current, can tell aluminum from gold, and pretty much any metal by the current drop. All metals have a different rate of current drop. Sorry, but I don't know if they are talking about decay time or what.

 

[Rudy]

Thanks Rudy,

I know they use the same system with the Eddy current in sewage plants to identify and separate the metals. I have seen several that specifically said they, using the Eddy current, can tell aluminum from gold, and pretty much any metal by the current drop. All metals have a different rate of current drop. Sorry, but I don't know if they are talking about decay time or what.

I don't believe that it is possible to separate the metals with a high degree of accuracy using Eddy currents.

 

[LatentDigger]

Way, way over my head. But interesting. Thanks. More reading on that somewhere?

 

[Tom_in_CA]

.......

I'm still stuck on the idea of a detector that can tell gold from aluminum.

............

If anyone ever steps forward (as many have done) to try to say that they can tell aluminum apart from alloyed gold (presuming coin/ring sized objects), or if they claim such a machine has been invented, then: Here's what you do :

Merely invite that person out to the nearest blighted inner city park, and turn them loose. See how well their "tones" and "sounds" and "mellow edges" etc.. . works. See what percentage of gold they're finding. And the percentage of aluminum they're leaving behind. Shucks, md'rs would even tolerate 50 to 1 odds/ratio, wouldn't we ?

But alas, you will never find anyone to accept this challenge. You will be met with the sound of crickets.

So alas : alloyed gold and aluminum share the same TID ranges, on a size-per-size basis. And the only technology we have tells only Conductivity, NOT composition.

 

[Diga]

What if were to forget about SMF and instead use SF's in a "differential" manner to tell the difference? For example:

A piece of aluminum is first hit at 2 KHz then 100 KHz. The difference between the returned signal at 2 KHz and the returned signal at 100 KHz is measured as "A".

A piece of gold is first hit at 2 KHz then 100 KHz. The difference between the returned signal at 2 KHz and the returned signal at 100 KHz is measured as "B".

Would there be a difference between A and B, and would it be enough to differentiate between aluminum and gold? OR, did I just kind of explain what SMF already does?

 

[Cherry Picker]

I don't believe that it is possible to separate the metals with a high degree of accuracy using Eddy currents.

To read on what sewage systems use to identify and separate metals is the same as a metal detector. At least it sounds like it. They claim to be able to use the Eddy current to identify all metals, and based on the "current drop". There must be some bit of accuracy if the sewage plants rely on it.

 

[Cherry Picker]

So alas : alloyed gold and aluminum share the same TID ranges, on a size-per-size basis. And the only technology we have tells only Conductivity, NOT composition.

I agree with the knowledge we use so far on metal detectors. The same type of transmitting & receiving, via Eddy current, at sewage plants can and do tell metal composition by the Eddy current. Or so they claim.

Interesting reading.
Eddy Current Separation

Electrical Conductivity Identification of Waste Particles Using Eddy-Current

Eddy current is also used for testing gold purity.

Eddy Current Testing for Authenticating Gold and Silver

 

[CountyWorker]

Thanks Rudy,

I know they use the same system with the Eddy current in sewage plants to identify and separate the metals. I have seen several that specifically said they, using the Eddy current, can tell aluminum from gold, and pretty much any metal by the current drop. All metals have a different rate of current drop. Sorry, but I don't know if they are talking about decay time or what.

Heavy metals in waste water are detected in a variety of ways. AFM, ASS, XRF, ICP-MS, etc. The food industry uses PI.

I remember reading something, about five years ago, about using lasers to penetrate soil for decontamination purposes. And then recently, I read an article about how gold excites uniquely from being bombarded with laser pulses.

Who knows, maybe our future lies in laser technology?

 

[Diga]

Who knows, maybe our future lies in laser technology?

That would be awesome. Plus, instead of dead round spots, we would get the more interesting...

 

[Tom_in_CA]

.......

I remember reading something, about five years ago, about using lasers to penetrate soil for decontamination purposes. And then recently, I read an article about how gold excites uniquely from being bombarded with laser pulses. .....

Westinghouse was working on something like this ^ ^ , for military purposes. Eg.: To tell explosives apart from other compositions, at-a-distance. Almost like a star trek tricorder sci-fi.

HOWEVER : It was mounted on a bobcat tractor (at the time I read about it in the early 1990s ). Cost millions upon millions . And you had to have scores of government clearances to use it. And you had to wear a lead suit (like a space suit) to operate it . And no one could be within a block of you.

I talked to one of the engineers on that ^ ^ project, and spelled out the distinction between an example pulltab, and an example gold ring WHICH COULD BE IDENTICAL IN EVERY REGARD with our current discrimination on metal detectors. And asked him : "Could your machine tell which one was the pulltab, and which was the alloyed gold ?". He said "yes".

But it was simply not worth it. Nor could it be "shrunk" (he said at that time), or avoid the dangers involved. So you'd be better off to simply dig up each target AND SIMPLY LOOK AT THE COTTON PICKIN' THING !

 

[IDXMonster]

There are certainly ways to up the odds of finding gold but the detector itself doesn’t come into play until the other “odds makers” are satisfied. I’m glad the conversation marches on about more efficient ways of finding gold though, it truly is the holy grail of our hobby.

 

[AccurateCalls]

There does not exist a bijective function that maps the set of all possible conductive objects to a TID / Set of impedance vectors given the number of distinct variables that define the set of all possible conductive objects. Very generally speaking this is the multi-variable induction problem at work. (This is a fancy way of saying, if somebody tells you the target produces an ID of 15, there is no way to know what that object is, since multiple objects could produce that TID of 15)

Corollary: Given a non-bijective mapping from an infinite set to an output space, it is always possible to find a finite subset such that a bijective mapping occurs.

Such a mapping function could exist within a highly restricted subset of the set of all possible conductive objects. E.g. if a field was populated exclusively by identically shaped and sized spheres that were made of either aluminum or gold, then the finite object space would preserve both injectivity and surjectivity. E.g. gold spheres would be given a distinct and consistent TID from aluminum spheres. So if the number of variables at play are restricted, then we do not suffer from the multi-variable induction problem.

If we assume the existence of such a detector that could distinguish all possible gold objects from all possible aluminum objects, then that detector would have to have a TID scale that ranges from 0 to infinity. E.g. it would be a detector that gives every distinct object a unique TID, and since the set of all possible conductive gold and aluminum objects is infinite, it would require an infinite range of TIDs.

It's possible in certain highly restricted scenarios for a metal detector / eddy current analysis to be able to reliably output the composition of the object (assuming enough variables are restricted for a bijective mapping to occur between TID and composition). I didn't read much into the paper on Eddy Current Separation for sewage, but it seems they are referring to metallurgical waste in which the conductive materials have effectively 'dissolved', thus possibly restricting their size and shape (e.g. if everything is a small particle of material), then it may be possible for them to determine the the concentrations of various metals in such waste.

 

[hoser]

I'm no electronics wizard or a metallurgist. and I do agree with Tom. To me there are just too many variables involved. Aluminum for instance, I found this description: "Aluminum is a metal that can be combined with other elements including copper, magnesium, silicon, zinc, and manganese, to alter its mechanical and physical qualities." Depending on which of these alloys are used it would change signal strength in my feeble knowledge base. So, just in case I'm not thinking straight, unless both metals, pure aluminum and pure gold are used I really don't believe anyone could definitely tell the difference without a shadow of a doubt. But that's just me.

 

[Carl-NC]

Would there be a difference between A and B, and would it be enough to differentiate between aluminum and gold? OR, did I just kind of explain what SMF already does?

Yes, that's what you did.

With only eddy currents alone, you can't distinguish aluminum & gold. If you also have the mass then you can go much further, and I suspect this is the case for waste separation.

 

[Diga]

Yes, that's what you did.

Damn it! I thought I was on to somethin'

 

[Tom_in_CA]

....Such a mapping function could exist within a highly restricted subset of the set of all possible conductive objects. ...........

If I'm not mistaken, your entire post is basically nothing more than what used to be called : "Ring enhancement programs". Because every since TID was introduced to the market (starting in 1982), it became no secret that you could play the "Las Vegas Odds". And notch IN certain zones, and notch OUT certain zones.

Because, for example, if you were to test a random 1,000 gold rings (like, let's say your friend owned a jewelry store, so you could spend an entire night air sampling each one). Let's say you made meticulous notes of each one's TID #. Then let's say you went to a typical junky blighted park, entered into "relic mindset mode" and dug : 1,000 aluminum junk items. Then you spent all night sampling the TID #s of these recurring typical aluminum junk park items.


THEN AFTER YOU'VE DONE THE ABOVE COMPUTER SPREAD SHEETS, then yes : You would indeed notice that there are TID # ranges where it's most likely to be trash . And other #'s with a higher degree of gold rings in that zone.

FOR EXAMPLE : Very few gold rings read down in the foil range. Unless you're talking very dainty ladies rings. Of which perhaps only 5% of the 1000 gold rings fall into that TID range. Contrast to 50% of the aluminum you dug might fall into that range. Hence presto: You knock out those ranges. And for example you knock out nickels' exact TID #, since : Odds are , it will be a nickel AND NOT GOLD (since perhaps only 5% of rings fell *exactly* on that #). And you would knock out beaver tail round tabs (of the most common recurring soft-drink). Because go figure, those are the bane of detecting. And perhaps only 20% of the gold rings fell in that range.


However, you open up all the other #'s, and perhaps that still accounts for 50% of the rings (since men's fatties read well above round tab). This was called "Ring Enhancement programs". And several wrote them (after these type studies), that users were supposed to program in to their Spectrum, XLT, Eagle, Teknetics, etc....


But the devil was in the details : YOU WILL STILL MISS GOLD RINGS AND YOU WILL STILL DIG A BOATLOAD OF TRASH. And if you're in an area where lawnmowers made can slaw out of aluminum cans, then : You can kiss these programs goodbye. You can chuck them out the window. Same for place where BBQ's go on (parks with campfires, BBQs, etc...) . Because those molten aluminum turds READ ALL OVER THE RANGE.

And sure as heck, someone's bound to try to claim that : That they can combine ring-enhancement programs , with sound characteristics (claiming gold is mellower or rounder or softer, etc....) . Here's what you do to solve this person's problem : You merely invite them out to the nearest blighted urban turfed park. Turn them loose. And : You will hear the sound of crickets. You will immediately cure them of their notion.

 

[Diga]

With only eddy currents alone, you can't distinguish aluminum & gold.

So that means the guys who claim they can tell aluminum from gold have managed to break the laws of physics! WOW!

 

[AccurateCalls]

If I'm not mistaken, your entire post is basically nothing more than what used to be called : "Ring enhancement programs".

Nothing in your reply has anything to do with what I posted. I never mentioned rings, programs, enhancements, nor ring enhancement programs, nor any of the other stuff you mention. Please don't construe this as me either agreeing nor disagreeing with you in either capacity. Perhaps I did a poor job explaining myself, or perhaps you read a few words of my post and then started typing your reply. I don't know. Here is my attempt to explain myself better.

Given a field populated by otherwise identical spheres of gold and aluminum, a metal detector could distinguish between the two with 100% accuracy. This is because TID would represent a bijective mapping in this scenario. e.g. All gold balls give TID of 20 and all aluminum balls give TID of 30)

Given a field populated with aluminum and gold objects with randomized properties (shape, size, orientation, etc), a metal detector could not distinguish between the two with any accuracy beyond randomly guessing. This is because TID would NOT represent a bijective mapping in this scenario.

The difference between those situations is the number of variables at play. The first situation has only one variable difference, the composition, while the second has multiple variables. In the first situation this allows for inductive reasoning to conclude what you are about to dig (e.g. all prior gold balls gave TID of 20, while all prior aluminum balls gave TID of 30, thus this target that is giving a TID of 20 that I have not yet dug, must also be a gold ball since this space is populated only by identical gold and aluminum balls). The second situation, one can not use inductive reasoning, given a TID, to conclude what object is in the ground, because there are too many variables at play. AKA The Multi-Variable Induction Problem.

Thus there may exist certain situations where there are fewer variables at play, or are known ahead of time, and thus a metal detector can reliably distinguish the composition of an object. E.g. in the Sewage scenario discussed above.