Bad Ground

[Qiaozhi]

Hi Mike,

don't use the microphone input. Use the stereo line input.

Microphone input is often restricted (band limitted) to the speech frequency range (max. 8 kHz). It has a high input impedance and has a high gain, which is very very noisy. Mic in is usually mono channel.

Well, some "incompetent hardware *******s" implement a low pass filter on a high quality high sampling rate sound cards (low-pass fc << nyquist frequency). I have one sound card, which has a SR of 96 kHz but the bloody ******* allows me only 20 kHz bandwidth (I can not see the 20-48 kHz range even if I sample with 96 kHz).

And we have some "incompetent software *******s", which don't allow the high SR (96 kHz). Some of them even can't make the damn software stable & reliable.

Not every sound card is convenient.
Aziz

Maybe it's because it's a "sound card" ??
The audio range is generally agreed to be 20Hz to 20kHz.

[Aziz]

Maybe it's because it's a "sound card" ??
The audio range is generally agreed to be 20Hz to 20kHz.

Well, a "good" sound card controls this via the equalizer function built in the hardware or software driver. They give the user the possibility of sampling extended bandwidth.

A sound card, which praises it's high 96 kHz sampling rate (SR) but does not allow sampling signals up to the nyquist frequency (SR/2) does not deserve it's high SR.

Aziz

[mikebg]

Measurement of H1(f) and H3(f) when you have a bucket bad ground

If the soil is conductive, its frequency response depends on diameter of excitation loop because it forms eddy currents and the value of fundamental timeconstant. For measurement of H1(f), the excitation loop should have the diameter of TX coil. For measurement of H3(f) the excitation loop should have the diameter of target (eddy loop diameter). If the target is a coin, the diameter of coin forms timeconstants of the ground for frequency response H3(f). That means the spectrum shape of H3(f) will be different from spectrum shape of H1(f), therefore we should make two measurements.

Let we measure responses when the target has size as a coin. The sensing system of measuring equipment consists of a plastic bucket filled with 150 mm soil layer as shown below. The bottom diameter of the bucket should be as diameter of TX coil. A coin size coil is placed on the soil. A large coil with diameter as TX loop is placed under the bottom of the bucket. If you have not enough soil, use a small bucket and place the large coil on the soil. The coin size coil should be placed under the bucket.

Measurement of frequency response H1(f)
For signal receiving we should connect the coin size coil as L2 in the circuit diagram. Resistor R3 is for damping. For excitation we should use as L1 the bottom size coil. If we use for measurement the soundcard of your PC, we have resistor R2 inside sound card. The resistor R1 is used as damping resistor. Capacitor C1 compensates the stray capacitance of the circuit. Note that the measurement needs a wide band circuit, therefore we should make damping of both windings.

Measurement of frequency response H3(f)
In this case for signal receiving, we should connect the bottom size coil as L2 in the attached circuit diagram. The resistor R3 has other value. For excitation we should use the coin size coil as L1. The values of C1 and R1 are different from previous case.

Before each measurement with ground, we should estimate frequency response of the system without bucket with ground. For this purpose, the coin size coil is placed on wood pieces to appear in 170 mm height from bottom size coil and measure the frequency response of system. Then we should reduce the height of coin size coil to 150 mm. The resulting difference of frequency response should have phase lead 90 deg relative to TX current for all frequencies.

[Dennis the Mennis]

For sale : worst ground ever

I just ran into a site wich should be the nightmare for any MetalDetector :

This ground is heavily polluted with the fruits of the Heinikus BottleCapus Ordinarus Tree :

If you want some of it for testpurposes : I'll send it to you free of charge , transport costs not included

kind regards ,

Dennis the Mennis

[simonbaker]

I just ran into a site wich should be the nightmare for any MetalDetector :

This ground is heavily polluted with the fruits of the Heinikus BottleCapus Ordinarus Tree :

If you want some of it for testpurposes : I'll send it to you free of charge , transport costs not included

kind regards ,

Dennis the Mennis

Ha! That is classic. Kind of a work of art.

It gives me an idea - we should use our finds to do market research on consumer drinking habits and become high-paid marketing consultants for various beverage companies...

[DOOLEY]

first thing i thought was "what a bloody mess"

if i saw that i'd switch detector off and walk else where.

to late now , many area's ruined by such rubbish , but has anyone invented a plastic
bottle cap yet ??

come to think of it , living in the modern age , plastic drinks "cans" with plastic ring pulls ,

and pref bio-degradable .

note: i'm not one of those tree hugging people

[Funfinder]

In Europe we have already alot beer in plastic bottles with plastic screw-caps.

I guess there is a chance to find something valuable at such a trashy place but first you have to clean up a bit.

Either use a very strong magnet, a rake or a large street-brush.


@ WM6
Don't tell uninformed persons out of the blue sky and without any reason that the Jeohunter would be comparable or just the same useful as your cheapo PI-selfbuilt kit for 50 bucks. You know nothing and you've been already fully disqualified by yourself!


@ MikeBG
Interresting ideas using the soundcard. But we would need more power. Perhaps the amp of a 100W subwoofer or something like this.
And if the higher frequency is to weak for penetration, we would need a stronger transmitter, too. If 100kHz can detect very small goldnuggets at the surface, those same 100kHz are for shure able to penetrate much deeper, if more power is used (as example 220Volts instead of 9 Volts).

[WM6]

@ WM6
Don't tell uninformed persons out of the blue sky and without any reason that the Jeohunter would be comparable or just the same useful as your cheapo PI-selfbuilt kit for 50 bucks. You know nothing and you've been already fully disqualified by yourself!

Which Jeohunter? What you are talking about?

[Funfinder]

Buy Jeohunter and you do not need to experiment anymore.

have you got the schematic ? .at that price i'd like someone to dig for me

You do not need schematic, there are only waste electronics inside.

bla bla bla *getting tired*

[WM6]

have you got the schematic ? .at that price i'd like someone to dig for me

bla bla bla *getting tired*

Dear friend, I am here for hobby and joy, not for building profit for others.

What are you doing last times? Some new inventions?

[satdaveuk]

imagine if you had to live there, must be a right old dump, but far as detecting it looks like tramps live there so i give the place a zero in finds

[mikebg]

@ MikeBG
Interresting ideas using the soundcard. But we would need more power. Perhaps the amp of a 100W subwoofer or something like this.
And if the higher frequency is to weak for penetration, we would need a stronger transmitter, too. If 100kHz can detect very small goldnuggets at the surface, those same 100kHz are for shure able to penetrate much deeper, if more power is used (as example 220Volts instead of 9 Volts).

Funfinder, the power of TX is not problem for a designer of metal detectors. We don't need more TX power, we need more competent design of block diagram.

My hobby is design of TX, RX and antennas for QRP amateur radio.
QRP method of radio communication means:
USE MINIMAL TX POWER!

The AIR & GND signal generated by TX is the main problem limiting sensitivity of metal detectors. A designer can increase TX power, but this increases the AIR & GND signal in input of RX. Then the designer should decrease the gain of preamp to avoid its saturation. The sensitivity of metal detector depends on gain of preamp (radio frequency amplification). Instead to think how to increase the TX power, the designer should think how to decrease AIR & GND signal in input and how to make RX preamp to operate always with maximal possible gain near to saturation of output.

Please see this:
http://www.geotech1.com/forums/showp...10&postcount=6
How changes the upper oscillogram when we increase TX power? This will decrease noise because we should decrease preamp gain to avoid saturation.
The competent designed metal detector should have the lower oscillogram. That means the sensitivity limited only by noise.

[WM6]

We don't need more TX power, we need more competent design of block diagram.

The AIR & GND signal generated by TX is the main problem limiting sensitivity of metal detectors. A designer can increase TX power, but this increases the AIR & GND signal in input of RX. Then the designer should decrease the gain of preamp to avoid its saturation. The sensitivity of metal detector depends on gain of preamp (radio frequency amplification). Instead to think how to increase the TX power, the designer should think how to decrease AIR & GND signal in input and how to make RX preamp to operate always with maximal possible gain near to saturation of output.

I support this approach and think it is a right way.

"Brute force" lead only to more problems not to more solutions (done many tests with PI in my testing plants in such matter, that I am aware - at least in world of mostly analog design).

Maybe some new high-tech uc solutions can solve some problems, but Aziz as our probably most skilled FFT etc. mathematicians here, do not aware me in this trough couple of years.

[simonbaker]

Funfinder, the power of TX is not problem for a designer of metal detectors. We don't need more TX power, we need more competent design of block diagram.

My hobby is design of TX, RX and antennas for QRP amateur radio.
QRP method of radio communication means:
USE MINIMAL TX POWER!

The AIR & GND signal generated by TX is the main problem limiting sensitivity of metal detectors. A designer can increase TX power, but this increases the AIR & GND signal in input of RX. Then the designer should decrease the gain of preamp to avoid its saturation. The sensitivity of metal detector depends on gain of preamp (radio frequency amplification). Instead to think how to increase the TX power, the designer should think how to decrease AIR & GND signal in input and how to make RX preamp to operate always with maximal possible gain near to saturation of output.

Please see this:
http://www.geotech1.com/forums/showp...10&postcount=6
How changes the upper oscillogram when we increase TX power? This will decrease noise because we should decrease preamp gain to avoid saturation.
The competent designed metal detector should have the lower oscillogram. That means the sensitivity limited only by noise.

I agree lower picture solves problem of "running out of overhead", i.e. saturating your front end with the null signal.

But once you have achieved lower picture, now you must ask where noise comes from. If it is circuit noise or external EMI such as power lines, then increasing TX power is a good idea, since you can decrease amplifier gain and therefore noise. If the noise is due to magnetic/conductive fluctuations in the soil, then extra TX power is probably useless. If you are out in the desert far from EMI with very uniform soil, then your main noise is probably circuit noise, and a high-power TX signal may be used to good advantage. But I agree, simply increasing TX power is not going to give magic improvements in depth without other smart engineering.

The most difficult challenge in the long run is clever processing to distinguish a target from soil fluctuations and trash.

At least that is how it seems to me. But I still need to make best TGSL I can, then worry about more advanced ideas!

-SB