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

[TheWizard]

Aziz,
thanks for LC hint...
I have not IB coil and first I need to find or make good coil...

My TX amplifier is full-bridge driver like Class D,
RX is Differential amp NE5534 x100

[Aziz]

Hi TheWizard and others,

you do not need a class D amplifier. Just use a simple low noise op-amp (e.g. NE5534 or better one NE5534A, AD797, LT1028 ), which delivers up to some mA driving current. Its completely enough. NE5534 can deliver several 10 mA on his output stage. So keeping the total number of active devices small, to achieve a better noise performance.

If someone does not want to drive the Sound card output, we can use a TX coil with a midpoint, which can be driven in differential mode with left and right channel. Right channel will be off phase of 180 degree. So getting twice more drive current and power to the TX coil.

Using good quality capacitors (polystyrene) in the LC circuit is tremendous important. Also the Q of the coil should be high enough, low eddy current causing wires (HF-Litz wire) and so on. Resonant capacitors should be placed directly nearby the coil and keeping therefore the cables as short as possible.

Once the TX coil is in resonance, the sound card is only delivering the loss of the resonant circuit and keeping the LC resonant swinging. With high Q of the resonant circuit, we do not need much power for this. The RX coil is a resonant filter, which only detects the TX signals. Other signals will be damped or filtered out. TX and RX coil should be shielded to lower the noise.

If we want to measure nano Volt changes from the coil, we have to tame the noise sources. Once done, there will be very surprising result.

Regards,
Aziz

[Aziz]

Hi Eugene,

I am just investigating the possibility of the digital VLF lockin MD. Once found all the important items, the next step will be of course using a DSP/µC board.
Using higher sampling rate with DSP's will allow a better phase seperation and SNR for the future. There are some good codecs available today: HD-quality, 24 Bit, etc.

It is only a matter of time. We just are making the proof of this type of MD. And it is allready done and it is working good and stable. Particularly the phase information of the Lock-in amp is quite good for easy discrimination.

There are much more tests to be done on the coming weeks and month. Lots of amps, LC-osc., coils, etc. must be tested yet.

Regards,
Aziz

[amtech2005]

Let's make a PC based metal detector with usb interface!!!

AZIZ, YOU ARE RIGHT !! # 1. IS "GETTING THE NANO-VOLTS OUT OF THE NOISE" AND # 2. WE WILL NEED PLENTY OF "THE CORRECT SIGNAL WAVEFORM "TO THE TX COIL SO SIGNAL IS DEEP , EVEN IN THE WORST OF CONDITION'S. WILL YOU USE A CODEC AUDIO ON YOUR DSP BOARD WHEN YOU RUN YOUR SIMULATIONS??? WITH DSP " SEPARATION AND ISOLATION OF ALL KIND'S OF WAVEFORM'S OR BANDWIDTH'S WILL MAKE LOCK-IN AMPS A REALITY. OR CAN THE LAP-TOP BE FAST ENOUGH ?????.........TAKE CARE............ EUGENE

[Aziz]

Hi Eugene,

AZIZ, YOU ARE RIGHT !! # 1. IS "GETTING THE NANO-VOLTS OUT OF THE NOISE" AND # 2. WE WILL NEED PLENTY OF "THE CORRECT SIGNAL WAVEFORM "TO THE TX COIL SO SIGNAL IS DEEP , EVEN IN THE WORST OF CONDITION'S. WILL YOU USE A CODEC AUDIO ON YOUR DSP BOARD WHEN YOU RUN YOUR SIMULATIONS??? WITH DSP " SEPARATION AND ISOLATION OF ALL KIND'S OF WAVEFORM'S OR BANDWIDTH'S WILL MAKE LOCK-IN AMPS A REALITY. OR CAN THE LAP-TOP BE FAST ENOUGH ?????.........TAKE CARE............ EUGENE

Laptops getting more attractive today. I can do almost everything with it, which DSP's can not or not with ease. Laptop has builtin such a HD (high definition) codec (sound card). And there is the possibilty to add-on a better sound card (I have the Sound Blaster Audigy 2 ZS Notebook with his excellent performance). Laptops delivers much more computation power than the DSP's. I can implement multiple lock-in amps nearby the detection frequency to get more information for instance. With 1 GHz speed, it can do the detection in real-time. With a sampling window of 1024 samples and 96 kHz SR, it is possible to make 94 measurements per second. With 512 sample window, the twice more. There is no problem of doing this and much more in real-time.

On the other side the DSP is smaller, has less weight and lower power consumption. But it is too difficult and expensive to build this. Who want to solder a 128 pin DSP chip? It's not easy. Also the PCB-board is quite expensive.

So first, I like the laptop detector idea. Everybody besides me has one at home. With some simple electronics, we can build a high definition digital VLF detector.
I will implement to whole code on it and if there is time and money, this can be a good option to adapt the solution to a portable one.

Regards,
Aziz

[amtech2005]

Let's make a PC based metal detector with usb interface!!!

AZIZ,YOU JUST GAVE ME ANOTHER IDEA. IN THE GROUND WHEN THERE IS INTERFERENCE , NOISE , MINERALIZATION , SALT, IRON. THE WORST POSSIBLE CONDITIONS YOU CAN THINK OF . THE NANO-VOLTAGE OR LESS, REACHING THE RECIEVE COIL CAN BE BOOSTED AND IDENTIFIED WITH 100% ACCURACY BY HIGH SPEED FREQUENCY , AMPLITUDE AND PULSE CHANGING CIRCUITS ,OR AUTOMATIC- HIGH SPEED SWITCHING DONE BY THE DSP OR BY HIGH-SPEED CPU ON YOUR LAP-TOP. IT WILL ADJUST ALL PARAMETERS OF THE TX COILS SO THE TARGET IN THE GROUND WILL BE CHARGED TO ITS MAXIMUM POTENTIAL, NO MORE FAINT OR ERRATIC SIGNALS. WE WILL CALL THIS "AUTOMATIC TARGET PULSING " AND THE -NEGATIVE OR + POSITIVE VDI NUMBER WHICH CAN BE CARRIED OR MEASURED TO THE THOUSANTHS ON YOUR FERROUS ,NON-FERROUS RECORDINGS[X-Y] AXIS. SO FOR INSTANCE A 36.726 READING CAN BE PURE GOLD AND A GOLD PLATED COIN WOULD READ 36.222. AND THE LAP TOP COULD TELL YOU PURE GOLD ."PLEASE DIG TARGET" OR "KEEP SWEEPING" ITS A BOTTLE CAP. JUST EXAMPLES................. AND THESE ARE STILL "DRAWING BOARD IDEAS". AND BEFORE WE REACH A CERTAIN LEVEL AND CLOSE TO COMPLETION WE WILL HAVE TO APPLY FOR A "PATENT" SO AGAIN WE ARE JUST TOUCHING THE TIP OF THE ICE BERG HERE . TAKE CARE...............EUGENE

[Aziz]

Just found an application note:

Noise Analysis In Operational Amplifier Circuits (Rev. B)
http://focus.ti.com/general/docs/tec...tName=slva043b

Pdf:
http://www.ti.com/litv/pdf/slva043b


TheWizard,

the AD797 delivers up to 50 mA at his output stage. So its enough for some basic tests. If this should not be enough, then a boosting transistor pair will solve this quite easy. The operating voltage is ±5V minimum.
The LT1028 has not enough collector current.

My INA128 is too much noisy. I will lower the input resistance and look the output on the coming days. But the NE5534 is working quite well and is good available.

Regards,
Aziz

[Tinkerer]

Digital Storage Oscilloscope

I have been using a PC based Digital Storage Oscilloscope to look at the various windows along the signal path of my PI experiments. It has helped me in many ways to see and understand what is happening to the signal. Specially for transients and ultra low frequency noise.
Time has come to look for a new solution for a low cost Digital storage Oscilloscope.
Can anybody recommend a low cost PC based Digital Storage Oscilloscope?
Tinkerer

[gwzd]

amtech2005,

I think - nano Volts (for Metal Detecting) is theory and dreams... in this range the noise is very strong

Finally! someone with a sense of reality!

Regards

[gwzd]

Just found an application note:


The LT1028 has not enough collector current.

An Opamp? collector current? What the heck is that?

My INA128 is too much noisy.

Lots of your noise comes from poor layout and lousy decoupling, check that first.

[Aziz]

Hi gwzd,

nice to hear from you.

An Opamp? collector current? What the heck is that?

Lots of your noise comes from poor layout and lousy decoupling, check that first.

What I meant was the max. output current of the op-amp.

I have on my instrumentation amplifier a second op-amp for the reference biasing in order to use in a single supply mode. Maybe the op-amp (OP177) is not the best choice for this. Second: the high input resistance of the amplifier. I have to test some configurations yet.

But the NE5534 is working quite well. It has internally lower input resistance. It delivers also quite enough power for driving the coils (some 10 mA). Two of them, one for RX and one for TX would be sufficient for a quite good working and cheap* detector (*without the laptop). I would like to test the AD797. It will probably work much better.

It is quite interesting to see the thermal and mechanical drift of the coil and electronics part. But this can be quite easily tamed with the software programming (with moving averages + hysterisis of standard noise deviation + some range). The result will be a stable detector.

Regards,
Aziz
PS: Now want to detect 1 Euro coin in 40 cm air.

[Aziz]

I can detect now lower than 1/100 degree of phase changes!!!
I want to measure the noise of my sound card tomorrow. This is critical and limiting device.
Aziz

[amtech2005]

Let'S make a PC-based metal detector with usb interface !!!

AZIZ , CHECK THIS LINK !! http://www.usbscope.eu// THEY ARE SELLING ON EBAY-UK FOR $32.00 LBS or EURO ?? Check it out , it may HAVE applications for the LAP-TOP DETECTOR !! ............EUGENE

[Aziz]

AZIZ , CHECK THIS LINK !! http://www.usbscope.eu// THEY ARE SELLING ON EBAY-UK FOR $32.00 LBS or EURO ?? Check it out , it may HAVE applications for the LAP-TOP DETECTOR !! ............EUGENE

Hi Eugene,

this hardware isn't good enough. Sound cards are much better.

Regards,
Aziz

[Qiaozhi]

I can detect now lower than 1/100 degree of phase changes!!!
I want to measure the noise of my sound card tomorrow. This is critical and limiting device.
Aziz

Such accuracy is totally useless in the real world. There are so many other factors, which affect the phase-shift returned from buried objects, that the ability to discriminate to even 1 degree accuracy is unrealistic.