Few general questin-opinions

[Davor]

Suhoj, there are a few misconceptions that you'll be better off without, so here it goes...

...which in turn cause standing waves...

No such thing in VLF or even PI detectors. They both operate strictly in near field conditions, so there is no matching, no waves or propagation, and reflection is very near 1. In terms of SWR we operate at SWR of 1:∞

LC oscillator with high stability and with low phase noise(no one has mentioned phase noise in TX osillator?!)

Because Rx is directly related to Tx phase, and all the phases are extracted directly from Tx, the very phase noise makes little difference unless it is related to amplitude noise as well. In case you are applying a phase noise infested exciter to a detuned Tx coil, it will result in PM to AM conversion, and there you have it. That's why many people tend to prefer low Q. (I don't)

We talking about levels of minus 60dB!!!

Actually we are talking about much less, but at about 60dB of isolation the preamp is not going to saturation, and is good to amplify signals some 60dB or so below that.

I am convinced that to day components are too good for kHz region!

That would depend mostly on your generosity and fullness of your wallet, and maybe to your wits. A typical Rx coil resistance is about 15ohm, resulting in ~0.5nV/sqrt(Hz) noise. Now, check Mouser or Farnell for op amps in 1nV/sqrt(Hz) range, and feel free to be amazed on the absence and price of components below 1nV/sqrt(Hz). The best integrated components of nowadays are merely copying the Cohen design that is over 30 years old.
Some 10 years back you'd be able to find much better components for noise than today because the consumer industry has no need for such components at the moment, and at that times you had MC and MM cartridges that needed them. Almost all real low noise components just vanished. Puff! Gone!

About factory made MD circuits I shall not waist my and yours time!I would only say,that I believe they far ahead in research than they put on the market!

Practically all the key people in MD R&D are contributing to this very forum. At least the good ones. With only a few exceptions, but judging from recent developments even they are lurking or trolling this forum.

What if we introduce negative feedback in RX front-and,in attempt to kill TX signal totally?

Follow daemon's posts, he does precisely that.

[Davor]

This is not true.

The double integrator oscillator is a circuit performing self sustained oscillations at a frequency determined by RC time constant of the integrators. With detuning it becomes less vigorous in oscillations, but it never stops overshooting or ringing. Every tuned circuit oscillator's amplitude response is delayed, and behaves exactly as an integrator, with slew rate and all. As you are dealing with the amplitude of oscillation, this analogy applies. Adding another integrator to the mix is a lethal combination. You'll find a very convincing example here http://www.geotech1.com/forums/showt...464#post168464
That's the power of double integrator in a control loop.

[suhoj]

Op amps are not only amplifier in real world,they are even not best ones!I would agree that op amps are good for lazy builder,because of relatively good stability with out too much efforts...
We can use tranzistor in variety of different kind as first stage in preamp and then use op amp...So,instead of using ready-made black box (IC),one can make it with discrete components in case he need it!If you wish,even HEMT are not expensive if you supply them from cheap SAT LNBC units...
Absence of standing waves in case of miss matching is news for me,I have no such experience in every day work,actually just opposite is proven truth for me.
Anyway,every one has right to think,but has no right to offend other people,in such philosophy I believe and strongly obey!
As Sinatra said,I shall do it "my way",if results are disappointing I shall publish them,if not-I shall publish it also,but by description.
Finally,I must admit-key for sensitivity is S/N ratio in every receiver,so MD RX is not exception,every rules and/or standards that apply for any receiving amplifier must be valid for MD too.Here also must be said that probably future of this kind of MD belongs to scalar waves because of theirs outstanding penetration abilities!

[Overtheedge]

suhoj, I think you might be looking at this from the RF engineering standpoint. My electronics background is in RF communications.

I had and still have some of the same ideas you have brought forth. But ...

I bought the book, "Inside the Metal Detector", and I'm starting to get the idea that this in NOT an RF type device. It is a transformer core analyzer. Several different ways of doing this, but the device tests the core (ground, etc.) for discontinuities by tracking hysteresis. Then through some sort of look-up table method, some MDs attempt to identify the identity of the discontinuity while determining the discontinuity's more exact location.

Effectively, these are not RF. There is no antenna. Just windings on a transformer minus the core. The frequency is chosen to meet the requirements of the discontinuity to be found and the hysteresis of the core material (the ground).

Hopefully I haven't mucked up the understanding even more.
eric

[Davor]

one has right to think,but has no right to offend other people

I'm sorry if you took it that way. it was not my intention.

[Carl-NC]

Anyway,every one has right to think,but has no right to offend other people,in such philosophy I believe and strongly obey!

No one is trying to offend, just pointing out that your thoughts may be appropriate for RF, but do not necessarily apply to metal detectors.

- We don't really care about phase noise, it is unlikely to ever get bad enough to matter.
- Force-driven TX oscillators are routinely drive off-resonance. It doesn't matter enough to matter.
- TX cross-talk into the RX results in DC offsets the in synchronous demods. We just calibrate it out and move on.

And so forth. Like you, my background includes RF design (GSM basestation chips) and I've had to shed some of those Biblical ways that would torpedo a GSM receiver if ignored.

[mikebg]

The double integrator oscillator is a circuit performing self sustained oscillations at a frequency determined by RC time constant of the integrators. With detuning it becomes less vigorous in oscillations, but it never stops overshooting or ringing. Every tuned circuit oscillator's amplitude response is delayed, and behaves exactly as an integrator, with slew rate and all. As you are dealing with the amplitude of oscillation, this analogy applies. Adding another integrator to the mix is a lethal combination. You'll find a very convincing example here http://www.geotech1.com/forums/showt...464#post168464
That's the power of double integrator in a control loop.

Davor, stray capacitances in every opamp form two or more integrators, but opamps operate excellent at correct design. Your circuit diagrams also can operate excellent if they are correct designed. See your "convincing example" redesigned:

http://www.geotech1.com/forums/showt...464#post168464