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

[TheWizard]

I will try to find C/Delphi API solution how to create and send in real-time, samples to sound-card... thats is the problem - precision RX/TX synchronization... all other is some math

Good way is external L&A but here problem is MCU with fast 16 bits ADC and fast PWM - this external L&A can have USB or RS232 and you can connected to Laptop, PocketPC even GSM(Java/Simbiam) mobile phone

[Aziz]

I will try to find C/Delphi API solution how to create and send in real-time, samples to sound-card... thats is the problem - precision RX/TX synchronization... all other is some math

Good way is external L&A but here problem is MCU with fast 16 bits ADC and fast PWM - this external L&A can have USB or RS232 and you can connected to Laptop, PocketPC even GSM(Java/Simbiam) mobile phone

I will try, whether L&A is possible in a good way using only HD sound card (96 kHz/24-Bit). It is interesting, how much the SNR will increase. This could take me several weeks because the experimental code must be revised and prepared to a good working product.
Aziz

[Aziz]

I found an interesting site with many schematics about amplifier suited probably for laptop MD or your ears:
http://sound.westhost.com/p-list.htm
Just check out.
Aziz

[vistac2000]

Hi AZIZ,

I select these circuits for TX and RX amp,
now i am ready for make a laptop detector,
please help me about this.
what kind of coil is good for our project? what diameter and what turn of wire need for rx and tx coils?
for Tx signal, we must use a signal generator software ok ?
for Rx signal, we must use a oscilloscope software or similar ok ?
i am waiting for your good ideas...
Thanks...

[Aziz]

Hi AZIZ,

I select these circuits for TX and RX amp,
now i am ready for make a laptop detector,
please help me about this.
what kind of coil is good for our project? what diameter and what turn of wire need for rx and tx coils?
for Tx signal, we must use a signal generator software ok ?
for Rx signal, we must use a oscilloscope software or similar ok ?
i am waiting for your good ideas...
Thanks...

Hardware:
Ok, your proposal should work anyway.

Coil:
Just use two equal dimension coils with impedance of 8 Ohms. If you like, then form a D type shape. Otherwise take the coils in circle shape. The balancing is quite easy for this type of coils. Small overlapping area of the coils should bring the coils in balanced position. The exact position should be seen on the software graphics output of the time-domain RX-signal. Diameter of the coils: any (let say 15-20 cm). Number of turns: any (let say 50-100 windings).

Software:
You do not need any signal generator, oscilloscope nor spectrum analyzer for this. The software should imply such functionality for the purpose of development and testing.
You should start read the Win32-API: Multithreading, Multimedia-API for sound card interface, GDI for graphics output, etc.
If you are not familiar with Win32-programming, then forget it. I can not help you with that. You have to write all the functionality by yourself. Also the whole MD software. I would suggest, you should first implement the sound card interface: just sending a pure sine wave form + acquiring the input signals and displaying them on the screen. For the purpose of stable display, perform a simple zero-crossing detector on the received signal (-> implementing oscilloscope). Then performing a simple FFT and displaying the power and phase spectrum (-> implementing spectrum analyzer).

One step by another.
Aziz

[TheWizard]

....Then performing a simple FFT and displaying the power and phase spectrum .....

NOT need FFT....

Soft is easy -
Create SIN TX buffer for sound card
Allocate RX buffer for sound card
Play/Rec TX/RX
Stop
Zero crosing and find 90 deg samples to calculate Amplitude and Phase

[Aziz]

....Then performing a simple FFT and displaying the power and phase spectrum .....

NOT need FFT....

You need it. Yes of course!
For the FFT filter later or for spectrum analyzing on multiple frequencies (disc).
Aziz

[Aziz]

By the way:
Doing FFT is also good for debugging purposes: detecting noise and disturbances.

I remember, I was searching for some strange signal former. I found this on the frequency spectrum. And it was my old analog monitor with his gigantic magnetic field disturbances. So you will see, where the spectrum is not disturbed.
;-)

[TheWizard]

You need it. Yes of course!
For the FFT filter later or for spectrum analyzing on multiple frequencies (disc).
Aziz

I speak about standart balansed, one frequency MD,
not for multiple frequencies...

And FFT is not magic method for metal detecting

[TheWizard]

The first test....

[Aziz]

Hi TheWizard,

nice App! Go on with development.

I found the very low noise AD797 and LT1028 Op-Amp. This could be used for the amplifier stage for the RX signals. I will buy one sample of these Op-Amps and test this.

[Unregistered]

Chookfoot

Hi,
I wish I knew more about this stuff, If you guys would just speak in a little simpler terms please.
What I can gather is that you are sending a typical PI signal out the sound card, through an amp then a coil. Then you are listening back through the coil, another amp then into the sound card.
Then using computer smarts to make some sense of the signal coming back.
Is that right ?

[TheWizard]

For PI metal detector, sound-card can not work - For Pulse-Induction you need ADC witht sample-hold up 500k samples per sec, sound card is only 44100, 48000, 96000 - too slow for PI....

[Aziz]

The sound card is used in a full-duplex mode. This means, while you sending a wave form through the transmit coil, you acquire in the same time the received signals from the receive coil. The acquired signals have a time delay (latency time). If you know the latency time (doing a synch), then you see the difference: transmitted signal causes some effects, which can be seen on the received signal. On induction balanced search coils, you will see a phase shift and amplitude change on targets.
The laptop is able to send any arbitrary wave form. It is possible, to search on multiple frequencies at the same time so doing a better discrimination.

Analysing on one frequency, the search coil can be used in resonant mode. This mode gives a better sensitivity and stability due to following facts:
1. on multiple radiants of the wave form the effects are summed.
2. resonant mode means, the coils are connected to a capacitor and they acting as a very sharp notch filter. Disturbing EMF is omitted and only the operating frequency is observed. On the resonant frequency, the coils have only the real part of the complex resistance: Z = R + j(wL - 1/wC). The imaginary part of the complex resistance Z gets zero on resonant frequency.
3. Eddy current effects can be detected much better.

Due to limited sampling rate of the sound card (48 or 96 kHz), only VLF/TR/Off-Resonance type detectors can be made. On the other hand, the ADC/DAC of the sound card is getting more attractive: 24-Bit and the high SNR (signal-to-noise-ratio).

That's all.

[Aziz]

I mean the inverted notch-filter! As notch filter acts as a band-stop filter, the inverted one is what I meant (band pass filter).

Sorry for this.

http://en.wikipedia.org/wiki/Notch_filter