Let's made a PC-base metal detector with usb interface !!!

[Aziz]

Hi Eugene,

AZIZ , WILL YOU USE 2 CHANNEL SOUND CARD AND THE USB ALSO ??? WILL YOUR SOFTWARE SHOW SPECTRUM ANALYSIS?? JUST WONDERING ? TAKE CARE . EUGENE

I will make the software (and hardware) as flexible as possible. So using one or two sound cards will be possible for different modes. No USB interface will be used. Only sound cards. USB interface is probably used for the power supply.

Simple one (single frequency):
one sound card, one LC-oscillator, using mic or line-in for RX signals, sound output for audio output (beeping). The software can detect the oscillating frequency automatically and the lock-in amp will do on this frequency.

Complex one (multiple frequency):
two sound cards, one for digital signal generation, using mic or line-in for RX signals, second sound card output for audio output (beeping).
This one will perform a multi-frequency search using multiple channel lock-in amps.

I will start with the first simple one: very easy

As the MD software is in experimental state, it will comprise: oscilloscope, spectrum analyser, lock-in amplifier, digital filter, some measurement tools (frequency measurements, inductivity measuremens, etc.). For maximum measurement throughput, these features only displayed on the demand.

Now looking for some convenient op-amps for the amplifiers.

Aziz

[amtech2005]

Lets make a PC-base metal detector with usb interface !!!

AZIZ, I Found a interesting manual for Ground imaging and the unit sell's for $6,000.00 us dollars and it's $1000 less if you use your own laptop. Anyway the pdf file is to big to post so I am trying to Break it down to post later. But anyway my question is How costly is a imaging circuit added to a High-Power Detector ??? I have a feeling the profit they make on image devices is Extremely High !!! Do Image devices use an X-Y axis type "Basics" . Like a old Graph Maker , x-vertical y- horizontal? Or maybe the combination of Ground resistance and magnetic differences. I know nothing about imaging , but do they really have a device that can see a silver dollar clearly at 6 inches deep or a gold ring ??? Or is it just a Red Blur on the screen ? Take care ............Eugene

[Aziz]

Hi Eugene,

I am not familiar with GI / GPR techniques. It is quite non trivial and I had no experiences with them yet. However, the laptop MD can detect all ground anomalies and can do a image map. I have implemented such a visualisation on the frequency domain (like a real-time radar display).


I have built a new amplifier for the laptop MD. It uses an instrumentation amplifier (INA128 ) and can be used with only single supply voltage (+5 V). It is better than conventional op-amp amplifier. I will replace this, when I can get the op-amp AD797 (ultra low noise, < 3 nV/sqrt(Hz)).

To make the laptop MD as easy as possible, I will implement only with the built-in sound card of the laptop. Using only line-input/mic for signal acquisition and stereo output for audio output detection sound.

A simple +5V LC oscillator which delivers enough power for the coil, will feed the balanced search coil.

+5V can be used directly from the USB connector of the laptop. So keeping all the hardware as easy and small as possible. All the other part will be implemented in software.

Yesterday, I have tested the amplifier and the result is amazing.

Aziz

[gwzd]

Hi Eugene,
... the laptop MD can detect all ground anomalies and can do a image map. I have implemented such a visualisation on the frequency domain (like a real-time radar display).
...

All ground anomalies?!! Kidding right?
Regards

[Aziz]

All ground anomalies?!! Kidding right?
Regards

Of course not. All possible detectable ground anomalies within the measurement range. Doing a multi-frequency spectrum analysis, all different ground types has his unique finger print. Also the type of target. I could show some spectrum responses in the future, if I have some more time.

For a good discrimination, it is enough to analyse two different frequencies. The more, the better the target identification.


BTW,
I built the CMOS inverter oscillator today (see on other topic) and it delivers good enough performance by its simplicity. It is not as stable as I want, however, in sophisticated software technique, this can be handled to meet good performance. It delivers much magnetic field strength.

The CMOS inverter oscillator (using CD74HC04E):
http://www.discovercircuits.com/DJ-C...kpowerosc0.htm

Topic:
https://www.geotech1.com/thuntings/s...ad.php?t=14222

I have operated the LC oscillator at +5 V and 7.5 kHz and the signal was quite strength enough to lower the rx amplification factor (amp ca. 10-20 times).

I will test some other LC oscillators (LC osc. with op-amp, colpitts, off-resonance, pulsed LC osc.) to get a reference platform for laptop MD. Also some coils must be still developed for this purpose.

Regards,
Aziz

[gwzd]

Of course not.

As I thought

All possible detectable ground anomalies within the measurement range.

With a metal detector? Still kidding, right? What about a concrete pipe without rebars buried 1.5 meters from the surface, do you really believe you can detect that anomaly with a metal detector?

Doing a multi-frequency spectrum analysis, all different ground types has his unique finger print.

Soils have three major parameters of concern, dielectric constant or permittivity (usually measured in relative units), electric conductivity and permeability (usually measure in relative units as well).

Also the type of target. I could show some spectrum responses in the future, if I have some more time.

Non metallic objects do not have measurable responses no matter how much lines of code you put.

I will test some other LC oscillators (LC osc. with op-amp, colpitts, off-resonance, pulsed LC osc.) to get a reference platform for laptop MD.

Aziz, you seem to be eager for knowledge and it appears you are learning a bit. So, take my advice, if you want stability then try with a PLL. They are not expensive to build and you will get much better stability than with all the other circuits you mentioned.

I wish you well,
Regards

[Aziz]

Hi gwzd,

I wish you well too. You should do some tests by your own.

[gwzd]

Hi gwzd,

I wish you well too. You should do some tests by your own.

Should I? On what exactly? We have 6 GPR units, 11 antennas, 5 EM profilers, 2 ABEM terrameters, 2 vibralocks a few no name metal detectors. where should I start Aziz?
Ah.. by the way, now the spring has finally come to us, 5 degrees Celsius outside today. It feels like Mallorca!
Have a nice evening

[Aziz]

Should I? On what exactly? We have 6 GPR units, 11 antennas, 5 EM profilers, 2 ABEM terrameters, 2 vibralocks a few no name metal detectors. where should I start Aziz?
Ah.. by the way, now the spring has finally come to us, 5 degrees Celsius outside today. It feels like Mallorca!
Have a nice evening

Hi gwzd,

you are equipped very well. Have you one no name MD for me? You know, I am a very poor man, who feels like Mc Guyver.
But software developing costs me nothing - I have all the time to figure out what I need.

Too cold for me. We had last week very high temperatures here. Today, it is raining a bit.

Regards,
Aziz

PS: Just looking for some LC oscillators..

[gwzd]

Hi gwzd,

you are equipped very well.

Not me personally but our company, and they are used very often. Not the metal detectors though. We work mainly with low enforcement, geophysical agencies and construction companies.

Have you one no name MD for me?

No, I don't, but if you are good at writing code and less expensive than the fellows in Hungary we usually use, then perhaps you can earn the money to buy a "named one".

You know, I am a very poor man, who feels like Mc Guyver.

I don't know who Mc Guyver is, but I assume very poor, I guess.

But software developing costs me nothing - I have all the time to figure out what I need.

It cost you time and that is the most precious commodity you posses. I assume you are young and therefore the optimism about time. Wait until you get a bit older.

Too cold for me.

for me too, but after you get roots in one place it is quite difficult to change even if it is a freezer like this northern part of Sweden.

We had last week very high temperatures here. Today, it is raining a bit.

Lucky you, we got soft snow which melts directly on the ground but none the less stupid snow.

PS: Just looking for some LC oscillators..

LC oscillators, Colpitts or Hartley which is the same dog with a different collar depending on where you put the L's and the C's. But as I said somewhere above these are extremely hard to master with good temperature stability, they can drift away heavily. So, take my advice and go for a PLL based design instead. More components? Yes. Less headaches? You bet!
Regards

[Aziz]

Hi gwzd,

I found regarding the last post interesting news from Eric Foster:
http://www.findmall.com/read.php?34,729335

and
http://www.geophyz.com/

As he mentioned, that different ground types can be detected by doing a multi-frequency analysis. This can be seen on my laptop MD, when I apply a band limited white noise to search coil and detect the spectrum response for the target/ground. This is quite interesting.

Regards,
Aziz

[gwzd]

Hi gwzd,

I found regarding the last post interesting news from Eric Foster:
http://www.findmall.com/read.php?34,729335

and
http://www.geophyz.com/

Thanks for the links Aziz, very much appreciated.

As he mentioned, that different ground types can be detected by doing a multi-frequency analysis.

It is not the ground what I'm interested in but the anomalies in it. For instance say you want to find a corpse in a yard 50x50 meters, perhaps is there or perhaps it isn't. Last year a young fellow went together with his snow scooter under the ice, what we needed to find was the body because the scooter was found pretty quickly. This kind of stuff you cannot do with an MD.

This can be seen on my laptop MD, when I apply a band limited white noise to search coil and detect the spectrum response for the target/ground. This is quite interesting.

I bet it is, but I'm more interested in well proven technologies that bring cash and MD is as the name implies for "metals". There are many other methods using coils which are not MD, for instance one of our EM instruments have four coils and they are using pretty much the same principles as the MD but when the conductivity of the soils drop (sandy areas for instance) then it becomes pretty useless. What I'm trying to say is that no method alone will give you all the answers.
Two years ago a whole "army" of fellows with metal detectors missed a large trunk with gold in Teselonika, just to be found easily by a guy using a GPR. The same, a large cave was not even in the radargram profile when a fellow using a resistivity meter found it. Just a couple of examples to illustrate the point. That's exactly why we do have a few MD, sometimes they are just invaluable and keep in mind that we are far from being experts using them.
All in all, I'm interested in anything that can show my customers what's below the surface with a degree of confidence. Yes, and divine rods too!!
If they can prove to me that 90% of the time their "guess work" is good then they are also welcome. No method will give you 100% confidence, keep that in mind. I've been in this business for a while and have seen quite a few that promise you the earth and the stars above just to miserably fail when the time of the truth comes.
I was in France in January visiting a customer and a fireman died while extinguishing a fire caused by a failure in the gas pipes. If they only had then a method for mapping all the gas pipes this tragedy wouldn't have ever happened at the first place.
Regards

[Aziz]

A little breakthrough in laptop MD with lock-in amp!

Hello Everybody,

I made a big step towards a really good working laptop MD utilizing lock-in amplifier.
Following condition:
LC-oscillator for TX coil
Instrumentation amplifier for RX signal
Coil not shielded, but operated on resonant frequency with capacitor connected (resonance filter). Coil not perfectly balanced.

Despite of lots of noise and instability of the LC osc., the lockin-amp can be slightly locked-in off the frequency, to get more stability. Also can not use the phase information, but the magnitude gets very stable and due to increase and decrease, the phase information can be taken from the magnitude. Never seen such a stable VLF detector with so much sensitivity!

There is much more potential yet:
a) more stable LC oscillators
b) more power through the coils
c) optimal coil (next step)
d) sophisticated software analysis

This is not a joke! Just make your own experiments, if you don't believe me.
Aziz

[Aziz]

Correction:

The slightly offsetting the lock-in frequency is the exact LC oscillating frequency. My FFT has not enough resolution to display the exact frequency and the lock-in amplifier shows me the exact frequency therefore. When locked to osc. frequency, the frequency error is at minimum and the MD is very very stable. Any small changes in the RX signal from nV to µV can be detected!

I can adjust the lock-in frequency in 0.1 Hz steps and can see the performance.

[Aziz]

Update for experimentators:

It has been shown, that the direct digital synthesis for the transmitting signal should be made directly in the sound card for better stability. This is of course internally crystal stabilized.
A better low noise small signal amplifier should be used. An Hi-Fi head-phone amplifier does the job perfect. It should use a low noise op-amp (e.g. NE5534A) with a line-driver circuit. A power of 100 mW or above is quite enough to keep the total distortion (klirrfactor) low.

The transmit coil is used in resonant mode (parallel with capacitor). The receive coil is also matched to transmit frequency (parallel with capacitor). The stability and the sensivity will amaze you!

First quick & dirty test:
1 Euro coin detected at 30-32 cm! (worst case coil, noisy environment, experimental software, etc.).
Phase shift information for discrimination is available and is quite stable!

Now focusing on better head-phone amps.

Aziz