TineFPGA-GOLD Metal detector

[SaltyDog]

Testing targets with a ring shape will give misleading results, as the shape itself will enhance eddy currents. Test targets are normally ball shaped for consistency, and to eliminate differences caused by orientation.

I am happy for you to send me over a lump of gold to test with ..

[green]

"My thought, maybe correct maybe not. Skin effect causes the decay slope to be faster at start of decay. Time constant=L/R, R is higher at start of decay because current is just on skin. AWG19 is closer to straight line decay than AWG10(thicker wire)"

I would agree with that.

How would you describe your TX transient?
How many Amps before switch off? What is the slope of the switch off transient? di/dt, or Amp/u second
How does this compare with the TC of the AWG19 closed wire ring?
Have you tried an iron wire ring of similar thickness wire?

Tester has two Tx on options. Constant rate(20A/sec to 20,000A/sec)1A peak or Constant current(.5A peak with 20us ramp up time) https://www.geotech1.com/forums/show...255#post240255

Scope picture of Tx off with constant rate(160us on, 1A peak), picture with log V to show time constant(adjusted damping).

Coil TC(.22us), AWG19 open TC(2.8us), AWG19 closed TC(28us)

Haven't tried a iron wire ring

[Skippy]

"send me over a lump of gold to test with"

I read DetectorProspector forum, and see the gold found in NZ. One tenth of a gram is a typical 'lump', you should easily be able to afford the going rate to purchase some test samples.

[Tinkerer]

Tester has two Tx on options. Constant rate(20A/sec to 20,000A/sec)1A peak or Constant current(.5A peak with 20us ramp up time) https://www.geotech1.com/forums/show...255#post240255

Scope picture of Tx off with constant rate(160us on, 1A peak), picture with log V to show time constant(adjusted damping).

Coil TC(.22us), AWG19 open TC(2.8us), AWG19 closed TC(28us)

Haven't tried a iron wire ring

Thank you for the answer.
Let's see If I understand right. When you say TX, cc 5ms, I understand that your TX pulse ramps up for 20us, to 0.5Amps, then maintains the current constant for 5ms (-20us) then drops to about 0A in 1.8us. (It looks as if there is some avalanche on the mosfet).
So I understand this as 0.5A/1.8us= about 0.277A/us.
To get the ampere-turns for the field strength, we would need to know how many turns on the TX coil and the diameter of the TX coil.

[green]

Thank you for the answer.
Let's see If I understand right. When you say TX, cc 5ms, I understand that your TX pulse ramps up for 20us, to 0.5Amps, then maintains the current constant for 5ms (-20us) then drops to about 0A in 1.8us. (It looks as if there is some avalanche on the mosfet).
So I understand this as 0.5A/1.8us= about 0.277A/us.
To get the ampere-turns for the field strength, we would need to know how many turns on the TX coil and the diameter of the TX coil.

The coil Tx above was done with constant rate 160us on time with 1A peak. The coil is fast so mosfet does avalanche at about 560V. Not sure what you are calculating, I'm thinking di/dt=E/L. 1.18A/us at avalanche or 2.1ma/us at 1 coil Volt. Only time decay rate is constant is at avalanche. Maybe wrong?

[SaltyDog]

An update:

With the latest tweaks, I have managed to isolate stainless steel from gold and silver as well .... so currently I have.

LED1: Stainless STeel
LED2: Gold/Silver
LED3: Aluminium
LED4: Nickel coin etc
LED5: Iron
LED6: Lead
LED7: Magnets

Copper is still not detected, I need to try some other forms of copper ..

[6666]

What are your plans for selling your kit.
Have you tried using coils made from different types of wire that is easily obtainable, such as enamel covered or pvc covered multi strand tin plated copper as the cotton covered wire is a rare item these days

[Carl-NC]

There are numerous alloys of stainless steel, some completely non-ferrous in response and overlapping low-conductive gold. Aluminum can be all over the conductivity range depending on size/shape/thickness, just like jewelry and nuggets. Not sure how you can definitively say "this is aluminum" and "this is gold." I expect as you expand your collection of test targets you will find these categories aren't so clear.

[SaltyDog]

What are your plans for selling your kit.
Have you tried using coils made from different types of wire that is easily obtainable, such as enamel covered or pvc covered multi strand tin plated copper as the cotton covered wire is a rare item these days

Regarding selling a kit:

I intend providing a kit somehow (a first for me) that includes a programmed TinyFPGA module complete with it's carrier board ... all parts populated ..
All you will have to do is add coil & case etc ..

I have tried 3 standard coils including two white's coils, all perform well ... my auto-calibrate function takes care off differences in coil/battery voltage ..
(Spider weave will be better for small gold though ... faster)

[6666]

Regarding selling a kit:

I intend providing a kit somehow (a first for me) that includes a programmed TinyFPGA module complete with it's carrier board ... all parts populated ..
All you will have to do is add coil & case etc ..

I have tried 3 standard coils including two white's coils, all perform well ... my auto-calibrate function takes care off differences in coil/battery voltage ..
(Spider weave will be better for small gold though ... faster)

[SaltyDog]

There are numerous alloys of stainless steel, some completely non-ferrous in response and overlapping low-conductive gold. Aluminum can be all over the conductivity range depending on size/shape/thickness, just like jewelry and nuggets. Not sure how you can definitively say "this is aluminum" and "this is gold." I expect as you expand your collection of test targets you will find these categories aren't so clear.

I have tried about 5 different Aluminium items, including pull-tabs, Al cans, Al bottle tops and Al sheets ... they ALL come under the correct LED ... I will test others as well of course ..

[Carl-NC]

I make test targets out of aluminum foil, each is 1" square with thicknesses of 1x, 2x, etc up to 32x.



On a VDI machine they progressively read foil up to about a zinc cent. In theory you could keep increasing the thickness and match silver coins. In fact, I have an aluminum token that exactly matches a silver dollar on a V3.

And if you pound silver thin enough, it will read as aluminum. Point is, with eddy responses any metal can mimic any other metal.

[SaltyDog]

Another test I did:

Detecting nails in the floor boards (on end) at 10" away, and determined they are iron ... probably rusty ... house age is from the '50s

[SaltyDog]

I make test targets out of aluminum foil, each is 1" square with thicknesses of 1x, 2x, etc up to 32x.

[ATTACH]49916[/ATTACH]

On a VDI machine they progressively read foil up to about a zinc cent. In theory you could keep increasing the thickness and match silver coins. In fact, I have an aluminum token that exactly matches a silver dollar on a V3.

And if you pound silver thin enough, it will read as aluminum. Point is, with eddy responses any metal can mimic any other metal.

Appreciate the comments. However they are "general" comments & as you say "theory", it's just not how my unit reacts "in practice" ...
.... and this is not a VDI machine, totally different machine & should not be used for comparison .... my 2c

[SaltyDog]

Another test I did:

Detecting nails in the floor boards (on end) at 10" away, and determined they are iron ... probably rusty ... house age is from the '50s

Another interesting test, is regarding EF (Earth field). In practice I see no response from swinging my detector head at maximum sensitivity.
I do get a response when placing a large horse-shoe magnet next to the coil .... which is how I derived the LED number for it ..