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

[Aziz]

Surprise, surprise!

Well,

I took a P-mosfet, that I have at home currently. The max. flyback coil voltage is approx. -100 V (IRF9540). If I get the IRF9640 soon, it will decrease down to -200 V.

The basic principle is working with direct sampling (no S&H, pre-amp fed into sound card directly). In the basic configuration (not optimized yet), I can detect a 1 EUR coin at 30 cm (coil diameter: 20.5 cm, mono coil, air test).

Not bad to continue with experimenting.

Aziz

[Aziz]

Postman! Where are you? Where is my packet?

Still waiting for the new parts (particularly for IRF9640 and BYV28-200). Coil driver finished. Pre-amps finished.

The software has to be extended. As I have no sampling timing control in the new passive board, except for cycle period and pulse width (no MCU), two pre-amps will be driven with different gain and will be operated in saturation mode. The difference time from saturation recovery of both pre-amp outputs will be measured in the laptop. The gain of the pre-amps control the first and second sample timings. The output of the pre-amps will be fed directly to the sound card inputs (capacitive coupling of course).

Let's see, what comes out.

Aziz

[simonbaker]

Hello friends,

I am quite an unlucky guy last days. The parts I have ordered last week gone back! The postman did not find my door.

Nevertheless, I have built as far as it was possible:
- Power supply finished
- DC/DC converter finished
- Clock generation finished (variable pulse width control through left and right channel of the sound card output)

The timings are precise enough. Only +/-0.001 degree phase noise at 375 PPS (Hz).

My oscilloscope is doing a really good job. I should have had it much earlier. See below my oscilloscope and the internals of the DC/DC converter synchronisation. LTspice is doing also a good job.

So just waiting for the new parts..

Aziz

Your experiences are like a soap opera, very entertaining.

Nice oscilloscope, we will expect even greater things from you now.

I have the 465, fairly similar. It looks like you have a small hex screw under each channel ac/gnd/dc slide switch - I wonder what that is for? Just mechanical or electrical trim?

(By the way, my channel 1 ac/gnd/dc slide switch seems to have come loose and I am nervous to open up to fix it. I am surprised because I have not used it that much).

Keep up the great work!

-SB

[Aziz]

Hello friends,

I am quite an unlucky guy last days. The parts I have ordered last week gone back! The postman did not find my door.

Nevertheless, I have built as far as it was possible:
- Power supply finished
- DC/DC converter finished
- Clock generation finished (variable pulse width control through left and right channel of the sound card output)

The timings are precise enough. Only +/-0.001 degree phase noise at 375 PPS (Hz).

My oscilloscope is doing a really good job. I should have had it much earlier. See below my oscilloscope and the internals of the DC/DC converter synchronisation. LTspice is doing also a good job.

So just waiting for the new parts..

Aziz

[Max]

My oscilloscope arrived today!
It was a good buy: 5 (!) probes included!

Some pots needs contact spray or be replaced. Overall a good buy for 151 EUR including shipment.

I am very happy to see the signals now.......

Aziz

Hi,
sure... if you don't see anything can't do much progresses... i hope you'll find it really useful, I'm sure you'll fix it.

Then 5 probes are nice add to the whole thing... good having extra available.

Kind regards,
Max

[simonbaker]

My oscilloscope arrived today!
It was a good buy: 5 (!) probes included!

Some pots needs contact spray or be replaced. Overall a good buy for 151 EUR including shipment.

I am very happy to see the signals now.......

Aziz

Let there be light...

[Aziz]

My oscilloscope arrived today!
It was a good buy: 5 (!) probes included!

Some pots needs contact spray or be replaced. Overall a good buy for 151 EUR including shipment.

I am very happy to see the signals now.......

Aziz

[Aziz]

This project continues...

I have some real practice in soldering now. Todo list for the coming days:
- yet another DC/DC converter number 3 (now converting into negative voltage)
- yet another clock generator number 4 (same as the last one)
- yet another coil driver number 2 (improved for switch-off and p-mosfet)
- yet another power supply number 3

I have modified the DC/DC converter synchronisation. A discrete hardware watch-dog is implemented to switch off the transmit coil on missing reference clocks (appr. 40 ms timeout).

This experiment is destined only for the items/questions:
- passive PI board only (no MCU) + sound card + laptop
- slow(er) coil damping
- high PPS rate (up to 12 kPPS)
- MONO/DD/balun/two TX coils
- higher inductivity coils
- direct & continuous sampling on sound-card inputs (no modulator used)
- frequency spectrum analysis of responses (two channels)

As the transmit clock signal isn't triggered via MCU, this should be quite accurate and stable now. I will have more current draw due to low dampening resistors to make the coil decay slower. On the other hand, a high PPS rate could be operated with fast coil decay (with more sample cycles to process). I should have now a better signal quality (low noise) due to less stages. Only a two channel pre-amp should be enough. The continuous sampling should allow a better digital filtering (no phase loss of correlated noise and distortions).

I will see, whether there is some benefit of this simple architecture. The circuit design is finished and I can start with soldering. If there is some benefit, then it makes sense to add a MCU for user interface (remote control of laptop software).

The oscilloscope will delay one week. The money I have sent, came back (wrong account number). So there is some time to solder the new passive PI board.

Aziz

[Aziz]

Hi friends,

I have compared the clock timings between the old (using the LM311) and the new board (using low noise op-amp).
The LM311 version generates 10x times more phase noise and 30x times more voltage (logic level) noise. So it was a good step to throw the LM311 out of the window.

The MCU generates also quite heavy phase noise to the clock timings. I suppose, the bad power supply design and reference potential (GND, -5V) is responsible for this. So I am not satisfied with this solution yet. The MCU has also a huge impact to the other circuits (RF EMI), which greatly increases the noise. I have to find the reason in the coming weeks.

I will continue with the simple and passive version in the mean time. As I previously mentioned, it will have:
- cycle frequency control (up to 12 kPPS)
- pulse width control
- alternating pulse width control (see Pooles Patent)
- synchronized DC/DC converter (for -5V)

The sound card output (left and right channel) will control the cycle frequency and pulse width. Using different reference frequencies, a multiple pulse width can be realized. The pulse width will be controlled via the phase lag of the two reference clocks.

Due to the pulse width control, I will not support the alternate damping anymore. A fixed (slow) damping will only be used. Due to the slow damping, there won't be a much high flyback voltage. So the p-mosfet IRF9640 (200 V) is convenient for this application. This makes grounding of sound card/laptop and PI board really easy (high side switching configuration, coil grounded).

A reference clock detector is necessary, not to fire the coil unintentionally for long time on missing reference clock signals. A hardware watchdog will therefore be necessary to enable the transmit signal. I have taken this solution, to make the clock signals more stable (very very critical) by letting the clock generator inputs unbiased (biased to ground only).

I am almost finished with the circuit design and will start with the prototype soon. You will have the results, schematics and spice simulation files as soon as possible, if it shows good results.

Aziz

[Aziz]

Hi Max,

Hi Aziz,
will you implement lock-in amplifier in it ?

What's the purpose of MCU, not just timing I think.

Do you made computations with it... e.g. time domain analysis of received signals ?

Kind regards,
Max

The purpose of the MCU is as follows:
- Serial communication interface to laptop (USB or RS232)
- GUI control (general user interface: push buttons, potentiometers, temp. sensor, ..)
- Software simulation of another CPU (device firmware processor)

The software CPU will control:
- Cycle control
- Timing (Transmit pulse)
- Hardware peripherals (PGA, integrator timings, S&H timings, ..)

The MCU won't do any computations. It won't have enough processing power anyway. It won't know, what it does really (due to CPU simulation). So there won't be any metal detecting/ground anomaly detecting firmware in it (no patent infringements ). The device firmware will be downloaded just in time. Then the board extracts two analog DC signals, which they then be modulated and fed into the sound card. The decoding and processing of the signals will only be performed by the laptop using the sound card.

There are different ways to decode the signals:
- Lock-in amplifier
- Goertzel algorithm
- FFT

As the sound card only accepts AC signals, the analog DC signals must be converted into AC signal (modulator does it). We have a two channel (stereo) line on the sound card. So the two signals can be almost fully digitized with 24 bit ADC resolution. It makes real µV resolution with high dynamic range processing possible.

That was the expensive variant with MCU (active version).

Now the cheap one without the MCU (passive version):
- simple hardware (only sound card interface)
- direct sampling (pre-amp output fed into sound card input)
- processing in the laptop only

I do not want do any math in the MCU. The laptop offers much more processing power and flexibility. If the passive version is possible, then we can say goodbye to noise.

So all my effort is to use the sound card and laptop for ground anomaly detecting/metal detecting.

Regards,
Aziz

[Max]

Hello friends,

I still don't know, how the slow damping, direct sampling will work. I am keen on to know this before I go on with the microcontroller version further.

The new reference clock generator seems to work really fine and delivers stable clocks. So this version will be used for the new simple and passive board (no microcontroller). I therefore will focus to this at the moment (I didn't forget to support you the schematics and spice files).

Now let's see, whether the very simple solution is possible.

Aziz

Hi Aziz,
will you implement lock-in amplifier in it ?

What's the purpose of MCU, not just timing I think.

Do you made computations with it... e.g. time domain analysis of received signals ?

Kind regards,
Max

[Aziz]

Hello friends,

I still don't know, how the slow damping, direct sampling will work. I am keen on to know this before I go on with the microcontroller version further.

The new reference clock generator seems to work really fine and delivers stable clocks. So this version will be used for the new simple and passive board (no microcontroller). I therefore will focus to this at the moment (I didn't forget to support you the schematics and spice files).

Now let's see, whether the very simple solution is possible.

Aziz

[Qiaozhi]

Hi Qiaozhi,



I have achieved the state of interfacing the sound-card with more precision now (less noise and less timing jitter). Two DC channel signals were modulated (converted to AC) to pass to PC through the sound card. The modulation frequency comes from the sound card (reference clock).

All development is done without any oscilloscope till now (still haven't one). But the sound-card scope allows me some precision measurements.

Next stage is generating the two DC signals from the front-end. The two DC signals will then be synchronously modulated (stereo modulator) and demodulated in the PC software for further processing. It allows a 24-bit ADC processing (provided that, the DC signals are clean enough). So small signal changes could be detected in the PC using different techniques (lock-in, boxcar integrator, ditigal filtering.. ) .

To develop the front-end, I urgently need an oscilloscope. I hope, it will come this week (bought one).

Aziz

I hope the oscilloscope purchase is successful. One of mine also came from eBay, and it turned out to be a good buy.

[Aziz]

Hi Qiaozhi,

Hi Aziz,

I don't yet understand where you are with this project.
Have you actually tested this circuit with some real targets yet?
If so, how does it perform? Or are you still in simulation land?

I have achieved the state of interfacing the sound-card with more precision now (less noise and less timing jitter). Two DC channel signals were modulated (converted to AC) to pass to PC through the sound card. The modulation frequency comes from the sound card (reference clock).

All development is done without any oscilloscope till now (still haven't one). But the sound-card scope allows me some precision measurements.

Next stage is generating the two DC signals from the front-end. The two DC signals will then be synchronously modulated (stereo modulator) and demodulated in the PC software for further processing. It allows a 24-bit ADC processing (provided that, the DC signals are clean enough). So small signal changes could be detected in the PC using different techniques (lock-in, boxcar integrator, ditigal filtering.. ) .

To develop the front-end, I urgently need an oscilloscope. I hope, it will come this week (bought one).

Aziz

[Qiaozhi]

Hi friends,

Although not all tests are finished yet, I will publish the schematics of the new board now. It consists of the following parts:
- Power supply, Reverse Battery Polarity Protection
- DC/DC Converter (synchronizeable)
- Clock Generation Modules
- Modulator Clock Divider (synchronizeable)
- Stereo Channel Modulator

The clock frequency of the DC/DC converter may differ to real circuit. So adjust the oscillator resistor and/or the capacitor for your requirements. The capacitor should be FKP2 or polystyrene one. A clock frequency between 30kHz to 100kHz should be ok. Resistors should be all metal film (1%).

The stereo channel modulator can be extended to modulate two channels more. In this case, the bipolar rectangle wave form modulation can not be used. But there is no need for this at the moment as I only want to use two independent channel processing. You can use NE5532/TL072 or compareable dual op-amps. There is also no need for JFET-input op-amp.

If you don't use the synchronisation of the modulator, then the polarity of the encoded channel signal cannot be determined except you feed a defined voltage to get the signal polarity.

Left channel of the sound-card output is reserved for cycle triggering and synchronisation (DC/DC converter and modulator). Right channel is for the modulator clock. The modulator clock will be either half or the fourth of the feeded clock frequency (jumper selection).

The new board makes the use of sound-card signal processing possible now. You only need to feed two DC signals into the modulator. The generated internal noise is acceptable and gives satisfactory signal decoding.

More to come..

Aziz

Hi Aziz,

I don't yet understand where you are with this project.
Have you actually tested this circuit with some real targets yet?
If so, how does it perform? Or are you still in simulation land?