SPI-EG4 !

[Monolith]

According Monolith, "Shielding the coil, this phenomena is attenuated or eliminated. In fact, I use it to test the shielding. If the detector does not detect the hand anymore, the shield is good enough."

According Charles Garrett, the shielding eliminates the "Green grass signal".

I think, the shielding can not suppress conductive signals because they are generated by eddy currents in substances shown below.

We can assume that the target response from the hand is caused by the body conductivity. The TX with the conductivity generates eddy currents.

But hen I tried an LRL rod, and hooked it (myself) up to the FFT, I got a very wide spectrum.

A similar spectrum like from the RX coil without the TX on. The RX coil's self resonance distorted the response somewhat.

That leaves us with more questions than answers.

Monolith

[6666]

maybe this one ?

[WM6]

We can assume that the target response from the hand is caused by the body conductivity. The TX with the conductivity generates eddy currents.

But hen I tried an LRL rod, and hooked it (myself) up to the FFT, I got a very wide spectrum.

Mysticism is only a point of view:

http://ezinearticles.com/?How-the-Hu...ies&id=1816572

http://boards.straightdope.com/sdmb/...d.php?t=239325

http://www.themetta.com/2009/11/how-...ansceives.html

[DOOLEY]

cheers 6666 for picture ,

why didn't people just say "bottle cap"

hello monolith ,

poss where i'm going wrong / or right , only testing the size of things i'm looking for ,
copper nickle coins , copper coins , gold rings ,
comparing with , ring pulls and screws and nails ,

so if i find a 10lb nugget , i might avoid it, if it responds like an iron pipe and not dig it up .

[Monolith]

Mysticism is only a point of view:

http://ezinearticles.com/?How-the-Hu...ies&id=1816572

http://boards.straightdope.com/sdmb/...d.php?t=239325

http://www.themetta.com/2009/11/how-...ansceives.html

Thanks for the links. Some interesting comments, but I prefer to avoid the mystical and stay with stuff that I can measure and reproduce.

Since I mentioned a LRL first myself, I better explain that.

When I first encountered the noise problem when using less than 10us delay with a PI,
I first looked at the signal spectrum that the coil captures, when it is used as an antenna. Above the coil's self resonating frequency the amplitude of the signal diminished, but the spectrum went way beyond.
Then I looked at the influence of my hands.
Then I thought, if my hands do that, what would an extension of the hand do? so I bent a 3.18mm aluminum TIG welding rod at right angle and held it in the hand like a LRL and looked at the signal spectrum. I posted the screen picture of that on the forum, but this is many years ago, I don't remember the thread or post.

Then I started to try many different shielding methods, until I found that simple verification of the shielding quality.

A long road that lead me to two simple rules that kept me happy until now.

The shield should not be detected by the detector.
The hand should not be detected by the shielded coil.

Nothing mystical.

But now I would like to know more about the mechanism of radiation or transmission from the body to wires or antennas or coils or RX boards.

What is happening?

Why is it happening?

How can I control what is happening!

Did I hear the word control freak?

Monolith

[Monolith]

cheers 6666 for picture ,

why didn't people just say "bottle cap"

hello monolith ,

poss where i'm going wrong / or right , only testing the size of things i'm looking for ,
copper nickle coins , copper coins , gold rings ,
comparing with , ring pulls and screws and nails ,

so if i find a 10lb nugget , i might avoid it, if it responds like an iron pipe and not dig it up .

Hi Dooley,

there is nothing wrong in testing common targets and rejects. I have maybe a hundred different test targets myself.
But I feel that if I classify the targets into a few categories it gets easier to predict the target response behavior.
Monolith

[Tec]

Target descriminiation PI

I like Mikebgs' pictures of grass and water and body ! there should be more picture posts.


I have been working on discrimination (Between different targets not iron discrimination) on this PI.


I think I am happy with the attached circuit .

The attached simple circuit can be used to tell the difference between targets:
Example:
5c, copper coin, bottle cap, 20c

Try it.

The more resistive a target, the more 'peaky' the signal.
Slower (conductive) targets decay slower.


On another note

If each target is like a 'fly wheel', some are light , some are heavy.
If you kick all the flywheels, the heavy ones will spin for longer, the light ones will spin quickly but stop quickly.

How they respond, also depends on the speed and weight of the kick.

I like this analogy, it helps me to understand PI I think.

Have you seen Newtons Cradle ?
http://en.wikipedia.org/wiki/Newton%27s_cradle

The way the Force (velocity and mass) is conserved :
One ball at one speed will be transferred to one ball at the same speed
(Not two balls at half speed)

Is this not similar to the way the change in magnetic field kicks the target.
What is important to note is that * faster change in field is not always better *

Slower decay will kick bigger targets better.
Faster decay will kick small targets.

So a particular mono coil will have a 'sweet spot', a target size that will respond the most.
Smaller and larger targets will have less response.

What do you say to that ?

i.e
a very fast decay curve on a copper coin is like kicking a tire with a bullet.
a very slow decay curve on a tiny gold nugget is like kicking spinning a top with a sledge hammer.

Thanks for reading this far !.

[Tec]

How I do the samples

The circuit hasn't changed much, and the main sampling technique is the best I can do for now..so I will post it.

I am getting sensitivity by building up multiple pulses on the capacitors which are charged with a constant current, after multiple pulses, I take a sample, then empty the capacitors.

So, 25 pulses, each pulse adds it's little bit to the cap.
A small change to the pulse width is multiplied 25 times in this case, the noise tends to cancel itself out..because it is random.

I don't dig bottle caps anymore, but pulltabs and rusty nails and bobby pins can still sound like coins.

I have been testing at the beach,
Only found a few $2 coins, and a couple of spoons...
And quite a few rust pieces,..and countless bobby pins.
I haven't found anything on the wet sand , only up where people sit...Am yet to find a gold ring..where are they ?
I read in 'gold gem and treasure', stories of hundreds of coins and digging 60cm for some of them..I don't believe these stories...lies

[Tec]

Android phone interface, and no gold

Wouldn't it be nice if recent posts were first, instead of buried at the end ?
That is my complaint out of the way

I now have an Android interface to the detector.
The phone is a Huwaei Ideos ($60), running android 2.1
I talk to it using a bluetooth interface (RN-42), these appear as a serial interface (uart) to your micro. (Less than $20).

So, the Metal detector is connected to the phone via a 114200 bps UART (with all the bluetooth stuff magically taken care of).

Now I can see the following on my phone (see screen shot): Signal level1, Signal level 2 and conductivity.
The two signal levels are the levels at two different sample delays, the conductivity is the ratio between them.

I hope to test it soon in the field. and give results..


Normal operation (locating targets) does not require the android phone,.. it is movement mode (must wave above targets) with a very fast response, can search fast, and takes no time (< seconds) for the audio to 'recover' after waving on a target.
I have this found to be the quickest and most sensitive algorithm.


No Gold yet..
I tried it at a creek, and found some nails and a magnetic rock (pic attached)...perhaps someone can tell me what it is ?,..it isn't an asteroid apparently. ?

Tec

[TH Jim]

Hi Tec,

Very interesting thread, particularly the discrimination. I will follow this closely.

Good work

Regards Jim.

[Tec]

A walk in the park (around a tree)

I have attached some results of detecting around a tree in an old park.
I'm sure lots of people have sat there over the last 80 years, and it seems every single one of them had a can of soft drink and threw away the pulltab.

I found 2 Florins ! (Pre-decimal Australin coins).

I have half of that tree left to go,... and 50 more trees in that park.

THe coin with >90% silver has a lower conductivity than the one with 50% silver,..this seems strange, but the metal detector doesn't lie !, so the alloy of 50% silver must have higher conductivity... though I haven't found any info to confirm that..

[DOOLEY]

hello tec ,

i allways like to search around trees , especialy big old ones , your right , during the life of the tree people cant resist climbing on and sitting under and enevitably loose all sorts of stuff out of there pockets.

found all sorts from coins still current (spent them!) to old 1914 penny's ,
nothing major though , oh yeh , and ring pulls as well.

[6666]

Any more news TEC ?

[Tec]

Update

Hi 6666,

The news is, I am selling the complete detector.
The design hasn't changed much just refined.

So:
PI circuit as per earlier posts.
Op-amp / comparator to convert signal into pulses.
The signal at two positions gives conductivity, different tones for different conductivities.

Biggest improvement has been shielding on the coil.
I have found Thin Al-foil with slits is just as effective as graphite powder on paper.
One CAN solder to al foil with lead free solder and flux.

The depth is a few centimeters better than a decent commercial VLF detector.

I am not going to post any more complete circuit diagrams because I don't want them to be copied.
The detector is called SPI-MX2

Discrimination by ear is possible, bobby pins and bottle caps sound obvious.

http://www.youtube.com/watch?v=VzPUn7kbVbI


Wire and complicated ferrous objects have complicated sounds, i.e double signals and different conductivities at different directions and heights.
Because they have two characters: magnetic field + eddy currents.

Non-ferrous sounds 'cleaner', i.e symmetrical when swept from different directions and one tone for conductivity.

Basically, I think it is a useful detector and competitive with what is already out there.
Not least because it is easy to use: i.e turn it on and start swinging.

..