Metal detecting with a computer

[gls]

First of all, sorry for my english, I'll try to express my thoughts the best that I can.

My idea is simple and probably "old" but I never seen a practical implementation, so I want to find a positive solution for a simple question:

"It is possible to make a metal detector out of a laptop, a coil and some little, common, simple electronics components?"

I my humble opinion this is possible. But what type of metal detector? I think BFO, at least for start, its principle is simple enough and most components can be emulated by the soundcard.

A soundcard can produce frequencies in the range o 50hz to 60khz (circa).
It is precise, adjustable in every parameter (waveform ecc.) and programmable.
Also, a soundcard can record imput in the same frequency range from the microphone line.

There are fantastic software out there for make spectrogram of the signal and "view" the target. Example: http://www.baudline.com/ This software also have a module for signal generation, useful for this purpose.

The problem is always the same: put a coil (or two) connected to the audio in/out. How?

Help me, please. This would be extremely simple to build, useful, flexible and... geeky cool

[gls]

Little precisation/hint:

Baudline can show at the same time in the same screen two graph: the one for the output generated, and the one for the input recorded (in realtime).
The two line of the signal overlap and only one of them is visible in normal conditions.
But if a frequency shift caused by a metallic target eventually occur, the two line will be visible. Maybe.
What do you think?

[Leto]

laptop detector

Hi gls,
have you seen folowing thread?
https://www.geotech1.com/thuntings/s...ghlight=laptop

low sampling speed of soundcards might be a problem, probably you should use best soundcard you can get..

[Max]

Hi,
for me soundcards aren't good for this kind of stuff.
Need a speedy hardware... e.g. some data-acquisition-card, you could also modify some existing one if you know where to find one suitable.

Also the idea of experimenting with BFO is no good. BFO are almost unused if not for educational purposes and kid toys...

You could made with a pair of cmos if want to try bfos... no need of laptop.

By me I'll try, instead, some Pulse Induction ... easy to trigger TX via e.g. the same hardware used for the acquisition... not so difficault, just need something that trigger e.g. an external device syncro pulse... then write some driver for the hardware. Not a big task for a motivated experimenter.

Yo could also use e.g. USB ports fast hardware interfaces to do same things.

In the worse case you could always trigger e.g. by parallel port a fast ADC then collect data... very easy.

Just my point of view.

Best regards,
Max

[gls]

Thanks for your answers Leto and Max

Problem 1

There are problems with soundcard because of low sampling speed... uhm...

A USB peripheral would be better. Ok.

This seem to me to complicate the construction of the interface, but if it is the only solution...
Sadly I cannot use the parallel port, my laptop do not equipe them . USB is better I guess anyway.

Where to start for this USB peripheral? Links useful?

Problem 2

BFO not very good. Ok.
Pulse induction however look difficult to me. VLF?

Current status: Idea for the project of "a sort" of USB VLF(PI?) detector.

[Max]

Thanks for your answers Leto and Max

Problem 1

There are problems with soundcard because of low sampling speed... uhm...

A USB peripheral would be better. Ok.

This seem to me to complicate the construction of the interface, but if it is the only solution...
Sadly I cannot use the parallel port, my laptop do not equipe them . USB is better I guess anyway.

Where to start for this USB peripheral? Links useful?

Problem 2

BFO not very good. Ok.
Pulse induction however look difficult to me. VLF?

Current status: Idea for the project of "a sort" of USB VLF(PI?) detector.

Hi,
there are a lot of ready to use USB to TTL interfaces (and also 3v etc) look on ebay. You can find a cheap one from e.g. china and then interfacing other parts.

There are also a number of projects , schematics on the Internet, few components and even ready to use softwares.

VLF on PC seems much complicated to be done.

Kind regards,
Max

[J_Player]

Pulse induction on pc seems easy because all signal TX and RX is time based. With a high enough speed interface, the computer can be used to set the main timing signal as well as for adjusting all the timing parameters. But the real advantage is that you can display the received signal on the screen like an oscilloscope rather than relying on a sound signal to identify targets. For advanced experimenters, the sampling could start during the transmit pulse to provide additional discriminating information.

The only advantage that comes to mind for using a laptop is to see the Rx waveform. It almost seems that the amount of work in setting up a high speed interface and writing the software for the GUI is more trouble than building a standard PI detector with an output interface to attach a small screen. If someone were to build a standard PI detector with an interface board that sampled the timing events along with the Rx signal, then fed these into a circuit to drive a small display, you would not need a laptop, but only the display and interface. The detector as well as the interface and screen could all be powered from the same battery.

Just some thoughts,
J_P

[Max]

Pulse induction on pc seems easy because all signal TX and RX is time based. With a high enough speed interface, the computer can be used to set the main timing signal as well as for adjusting all the timing parameters. But the real advantage is that you can display the received signal on the screen like an oscilloscope rather than relying on a sound signal to identify targets. For advanced experimenters, the sampling could start during the transmit pulse to provide additional discriminating information.

The only advantage that comes to mind for using a laptop is to see the Rx waveform. It almost seems that the amount of work in setting up a high speed interface and writing the software for the GUI is more trouble than building a standard PI detector with an output interface to attach a small screen. If someone were to build a standard PI detector with an interface board that sampled the timing events along with the Rx signal, then fed these into a circuit to drive a small display, you would not need a laptop, but only the display and interface. The detector as well as the interface and screen could all be powered from the same battery.

Just some thoughts,
J_P

Hi,
yes right. One could save much work just interfacing an e.g. tektronics lcd portable scope to an existing PI design.

What could be interesting making the stuff with a PC and some real time software is making something like in the Robert Holko's project... making e.g. filters and integrations just using routines instead of op. amps.

Much more flexible approach to PI problems cause you have to write just some code and don't rely on op. amp. specs and circuits. Just some fast ADC and software.

I don't belive is possible e.g. making something like that Holko did with uP-VLF , so disc, cause of problems related to the absence of a param like phase shift... in PIs. But could be interesting.
Problems are similar, needed really fast hardware for sampling.

Best regards,
Max

[gls]

Now I have a ton of article to read for this evening, but I begin to have a much clear idea
A "laptop approach" for me worth some study, even if it is not the best way.
Im really "obstinate"

PS: Thanks for your great advice friends!

[Sean_Goddard]

To do this you will need a sound card with AT LEAST a 96KHz sample rate and even then you will only be able to Tx areound 7KHz. No problems you could use a lower sample rate with a Tx of say 3KHz. Give it a try and see what you get.

[wyndham]

What CPU speed is needed for a program to work in compiled basic? That is the area I was wondering about. Wyndham

[Sean_Goddard]

Even a PIII 1GHz would do the trick, but a P4 dual core 1.66GHz would be overkill.

Why, and I'm going to give it ALL away here, Try a ADI Tiger SHARC or Blackfin DSP. More than enough power to do ANYTHING you want on the metal detecting front, problem is, the power requirements, a 12V 4Ah gel cell should do the trick though.

[wyndham]

I was thinking of a Pico atx board from Via that is 17cm x 17cm low power and one version has an onboard sound chip up to 1gig mem usb, 1 com port sound video ide and cf card. At around $200 might be worth looking into.
I only put the specs I remembered, only have several brain cells left and in this heat(NC) there's going fast.Wyndham

[Sean_Goddard]

Yup, that should do it, but you really need that faster sample rate and true 16 bits if possible or maybe switched gain amps on the front end.

Go for it, I have some ideas if you wish to PM me, failing that visit Pulsar Electronics and mail me I'll converse with you in private then if what we have works, you can post it here for all to use.