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

[Aziz]

Hi Mustafa,

Hi Aziz,
Thanks for all the information you shared with us. Keep up the good work.
Could you explain more about your power supply module?

P.S. Why are you still living in krautsland? Return to Turkey. I have a spare room if you want.

Battery: +11 .. +14.4V (common range)
Minus is connected to ground.

+5V digital (max 1A, using 7805):
MCU and digital logic power supply

+5V/-5V analog (max 50 mA, using 78L05 and 79L05):
Sensitive analog part and jfet/MUX switching logic. -5V is generated with the discrete charge pump already published here (voltage inverter). It is synchronized with the cycle clock signal.

+18V analog/digital (max 50 mA, using 78L18 ) :
Mosfet gate driver to lower the switch-on resistance of the mosfet.
Instrumentation amplifier for the fully differential PI front end (operated at +18V/-5V). As I am using N-MOS fets, I have a high common mode voltage at the coils (battery voltage). I need at least 3V more than the common mode voltage to operate the instrumentation amplifier with the N-MOS fet configuration.

If I am gonna to use P-MOS fets (less transmit power), then the instrumentation amplifier will be supplied with +5/-5V only.

+18V is generated with a discrete charge pump (voltage doubler), which is also synchronized to the cycle clock signal.

Aziz

[simonbaker]

Hi Mustafa,
Battery: +11 .. +14.4V (common range)
Minus is connected to ground.

+5V digital (max 1A, using 7805):
MCU and digital logic power supply

Aziz

It's too bad 7805 wastes so much power -- doesn't it?

Is there a way to use switching converter to step down more efficiently without a noise problem?

Regards,

-SB

[Aziz]

Hi simonbaker,

It's too bad 7805 wastes so much power -- doesn't it?

Is there a way to use switching converter to step down more efficiently without a noise problem?

Regards,

-SB

7805 is quite well suited even it powers up some efficiency (global warming?). The load regulation and noise performance could even be improved with a two stage voltage regulation: +12V -> +8V -> +5V.

Switching converters generate much noise. The more, if there is a SMPS implementation (inductive). The MCU won't consume much power (max 30-40 mA) and should not be an issue.

My power supply module consumes in the standby mode (no other modules) only approx. 20 mA @12 V at the moment.

We have much much more waste of power during the flyback phase. We should get this energy back and recycle it.

Aziz

[simonbaker]

Hi simonbaker,



7805 is quite well suited even it powers up some efficiency (global warming?). The load regulation and noise performance could even be improved with a two stage voltage regulation: +12V -> +8V -> +5V.

Switching converters generate much noise. The more, if there is a SMPS implementation (inductive). The MCU won't consume much power (max 30-40 mA) and should not be an issue.

My power supply module consumes in the standby mode (no other modules) only approx. 20 mA @12 V at the moment.

We have much much more waste of power during the flyback phase. We should get this energy back and recycle it.

Aziz

I see. I thought maybe you needed close to 1A at 5V - so not a problem!

Regards,

-SB

[Aziz]

Hi guys,

I made some little progress. Module number two is just finished and tested successfully. It is the sound card interface module. I had to add an automatic DC servo control to the modulators. Any offset voltage errors will be compensated and the integrators won't be saturated anymore to give a good dynamic range. The noise caused from the DC servo is really negligible. I have implemented a four channel modulator (each line is encoding two analog input signals) and the channel seperation of decoding is really good.

The decoding can resolve µV and even sub µV signals. The output of the modulator is very stable (magnitude and timing) and quiet (low noise).

I have to built now three more modules:
- mcu (processor, PC interface, user interface),
- front end (coil drivers, pre-amp) and
- signal processing module (integrators, S&H, .. )

As the modulators are working well, a four channel processing will be developed. This will give more flexibility to search modes.

Aziz

[Aziz]

Modulator CRO Output

Hi all,

the new channel modulator is working very well. The best modulator I have at the moment so far (cheap and easy).
Below is only one channel output (right line) with two analog signals encoded. Two different fixed analog signals were fed into the modulator. The modulator base frequency is 6 kHz. One analog signal channel is modulated with 6 kHz/2=3 kHz and the other one with 6 kHz/4=1.5 kHz. As long as the dynamic range of the input line is not exceeded, the channel seperation works well (change of one signal channel is not contributing to the other signal channel).

I have put the change of one decoded signal into the graph (violet color at the zero base line) and its FFT to see the spectrum (gold color). You can see a high frequency noise of the decoded signal, which can be elegantly removed by the software processing. Of course, any correlated noise can be removed. So there is no need for a specific 50/60 Hz filter at all.

Aziz

[Aziz]

Hi all,

I am just wiring the MCU module. It will have five interface connectors:
- RS232 serial port (using MAX232A or ADM232A, 120 .. 200 kbit/s).
- coil driver port (pulse logics, optional coil power control)
- signal processing port (PGA, integrator and S&H logics)
- ISP programming port
- and user interface port (remote control box)

I do not like the USB licensing politics nor I have enough money to buy a vendor and product ID for it. The communication to the laptop is not time critical at all and will be solved via USB-to-Serial adapters. It will make the development fast, cheap and easy and is working already well.

Remember, the laptop will just be powered up and stays in the backpack in idle mode (for least power consumption). The user interface will fully remote control the laptop detector software. No bells and whistles and no graphics output for detection operation (display will be switched off to save power). To set up the software, configuration dialog boxes will be implemented of course.

Most of the ADC inputs of the MCU (ADC0..ADC7) will be used for user interface purposes. Some of them are pure digital input ports for buttons. Mode and parameter selections will be implemented via potentiometers (cheap and easy).

I will use either the cheap 16 MHz ATmega16 or ATmega32 (40 PDIP versions). The speed is fast enough to control the hardware even there is a software simulated CPU in the MCU. The MCU will not do any calculations nor digital processing (hey, don't loos the objective! we have a much higher number crushing device: the laptop). If the MCU SRAM gets into its limits due to multi-cycle processing operations which require more device micro code, then the ATmega644 and ATmega1284 could be a option.

I will absolutely insist of a pure "firmware free device" for some reasons. I like the new modular development boards. I have enough space for some additional tryouts and experiments.

Aziz

[Aziz]

Hi all,

the development still continues but I am seriously missing a down-side level shifting IC!!! Like from 0/+5V logic into -5/+5V or 0/-5V logic. Do anybody know a good available integrated circuit level shifter for this application? Please, let me then know.

In case of there isn't a level shifter available, I have to use many transistors, resistors and schmitt triggers. Some IC's for S&H, analog MUX/DEMUX/Switches gives the best option for this problem: no level shifting required. This would make the design easy.

Some analog parts will require -5/+5V logic levels and some only 0/-5V. I will order some S&H ICs (LF398 ) to see their performance. I probably will try out 2-3 different signal processing modules.

The main architecture of this design will consist of two coil driver stages (for bipolar transmit pulses), one differential pre-amp, one PGA and two independent differential integrators (one for each audio line) with hold and reset capability for multi-cycle processing. On each output of the integrators, there will be following two seperate S&H stages. Total analog signal channels would be four then, which will be modulated from the modulators.

I have some fresh money to spend it for some new parts..

Aziz

[MTK]

This project of yours started out just fine. But now it seems too complicated.

KISS

Vapourware so to speak.

No offense.

[Aziz]

Hi MTK,

This project of yours started out just fine. But now it seems too complicated.

KISS

Vapourware so to speak.

No offense.

That's true. It isn't a KISS principle anymore.

I am not following a commercial product. It is destined to an experimenting platform to find new knowledge about this great stuff.
Just for fun.

Aziz

[jamal]

frined

hi aziz whare are you from


jamal

[Aziz]

Hi all,

I just got the ordered bunch of parts. Its not complete but I can go somewhat forward in the coming weeks.
I will try the worst case configuration with more integrated circuits using bilateral analog switches/muxes and S&H (LF398 ) .

Aziz



PS: Jamal, I am from Turkey. My passport is a kraut passport. So I am a so called "passport-german" (german: Passdeutscher). I am officially a kraut but seen and treated as a foreigner.

[Samy]

[quote = Aziz, 97886] Hallo alle,

Ich habe den bestellten Strauß von Teilen. Es ist nicht vollständig, aber ich kann etwas nach vorne gehen in den kommenden Wochen.
Ich werde das Worst-Case-Konfiguration versuchen, mit mehr integrierten Schaltungen unter Verwendung von bilateralen Analogschalter / muxes und S & H (LF39.

Aziz



PS: Jamal, ich bin aus der Türkei. Mein Pass ist ein Kraut Reisepass. Ich bin also eine so genannte "EU-Pass-Deutsch" (Deutsch: Passdeutscher). Ich bin offiziell ein Kraut aber gesehen, und als Ausländer behandelt werden. [/ Quote]




Slm. aziz güzel bir deyim hosuma gitti.

yaptigin isi merakla takip ediyorum.weiter so.!!!

[Aziz]

Hi all,

look at this: Viliv S7
http://www.myviliv.com/ces/main_s7.html

7 hours continious movie playback! Including a GPS and touch panel display. So a laptop MD with one day operation duration is getting more and more attractive now.

Aziz

[Aziz]

Hi all,

regarding the 24 bit HD sound cards and some comments on them:

There are good ones and bad ones. The two USB sound cards I already showed here (white and black one) differ quite.

They seldom achive 24 bit resolution due to heavy internal noises. You are quite lucky, if they even achieve 20 bit.
I have found, that my black sound card achieves 23 bits of resulution, if the input signal is attenuated as much as possible to reduce the noise of the sound cards internal amplifier. At 0 dB (no attenuation, 1Vrms input range), the resolution decreases down to 18 bits (noise floor at -110 dB).

The white one is much worser. The internal clock is not accurate. The black one is superior (very accurate clock and stability, lower noise). If the input signal is attenuated by 20 dB (x 0.1), the internal noise floor is at -130 dB (almost 22 bits). I even can set to a minimum input level (mixer ruler set to minimum possible), then the noise floor decreases down to -140 dB (23 bits). So the input signals should be rised in level from the external low noise circuit to use the full ADC resolution.

The big noise source is definitely the EMI (induced by the coil itself). So the conditioned four analog signals will be quite noisy. If they are somehow correlated, the software could compensate this (big advantage).

Aziz