the thinking of reverse for vlf

[lonelyWOLF]

hi.
my opinion...

if I get demodulator pulse from receiver for sampling of transmit signal . I get what result?. What do you think about this idea?

[mikebg]

Hi Wolf,
The disadvantage of conventional (upper) block diagram is that reference signal for synchronous demodulation and received signal have different phase instability. Reference voltage obtained from TX by wire is stabile, but the narrow band preamp in receiver provides instabile phase shift of received signal.

In the lower block diagram there is RF pulses for synchronisation made by shorten with electronic switch S of a winding in sensing head. The phase instability of narrow band preamp is not important because the signal and the reference voltage for synchronous demodulation are amplified by the same amplifier.
I think there is a patent for design solution shown with the lower block diagram.

[lonelyWOLF]

hi mikebg ,
Thank you for your reply. I'm not an expert about RF, many systems do research, I consistently do well and how can I sensitive the minimum phase shift. I see the bottom of your block diagram Five-number block .What is this? Why it wireless? we use always the multi-core cable. one wire use for switch pulse .Would you make a more explanation for five number block?..The answer to my question above. does not change anything, I have two signal one stable. other is shifted

best regards.

[mikebg]

Hi Wolf,
The abbreviation CW used for block 1 in posting #2 means “continous wave” ie TX radiates nonmodulated sinusoidal signal. As opposite to PI (pulse induction), the abbreviation VLF means sine induction.

The synchronous demodulator (block 4) operates as scalar multiplicator. It multiplies two voltages: received voltage and a reference voltage having the same frequency but with different phase. The demodulator operates as lock-in amplifier, ie it is phase sensitive and can suppress interference and noise. However synchronous demodulator operates not only as phase detector. Since it is also amplitude detector, to avoid amplitude demodulation of sinusoidal reference voltage, it is transformed in a square wave voltage having stabile amplitude. The output of block 5 is square wave.
Imagine that your radio receiver or TV set needs a very long wire connected to the transmit station. At this stupid principle operates every conventional metal detector according the upper block diagram in posting #2. The problem is not in the long wire. The wire is too short at metal detectors because the TX section is placed in the same box as RX section. The problem is that the wire connection delivers from TX a reference voltage for synchronous demodulation without phase noise and stabile phase. However in the other (the left one) input of synchronous demodulator, the preamp (block 3) delivers signal having phase noise and phase instabilty due to narrow band amplification in the front end of RX. The solution of problem is to obtain reference voltage from the same narrow band amplifier. This is not easy.

Every radio or TV receiver is wireless, ie no need of wire connection with TX to obtain reference voltage for demodulation, because every broadcasting transmitter radiates signal with carrier frequency. Can we use such principle for TWO BOX metal detector (without wire connection between TX and RX box)? The principle is shown in block diagram here below.

The RX box of a VLF MD receives without target two signals having carrier frequency: AIR signal having stabile amplitude and phase because is transfered directly from TX coil to RX coil, and GND signal, which amplitude and phase random change. Because of GND signal we not use the combination AIR&GND signal for reference voltage in a MD. This block diagram operates correctly for receiving of conventional FM broadcasting because block 5 is designed to operate with inertia, ie amplitude and phase of output can not change rapidly. However for metal detection even a slowly change of reference voltage means change of discrimination and elimination of GND signal (GND balance).
The block diagram shown below can operate as MD if the GND signal is suppressed by TWIN LOOP (second RX coil).

[simonbaker]

I think maybe you are both talking about different concepts. mikebg seems to be talking about reconstructing a sync pulse from the received signal so the direct TX signal is not needed for the sync pulse????

lonelyWOLF seems to be saying: why not use the RX signal for the sync pulse and "sample" the TX signal (the reverse of what we do in the TGS detectors).

In theory it (lonelyWOLF's idea) probably works and is worth keeping in mind as a creative alternative approach. My guess is that you lose the advantage of the Syncronous Detector's property of "integrating" out the noise in the RX signal. Instead, you are detecting phase by relying on the slight change in zero-crossing of the RX signal due to a target signal adding to the large null signal, which would then slightly shift the sync pulse and integrate a slightly different portion of the TX signal. The sensitivity therefore might depend on the size of the null signal, and noise occuring at the zero-crossing moment would be particularly significant.

But that is just a guess. Certainly an interesting and creative idea.

-SB

[mikebg]

I wish to upgrade the block diagram in posting #4 with suitable coil system: TWIN LOOP

[lonelyWOLF]

lonelyWOLF seems to be saying: why not use the RX signal for the sync pulse and "sample" the TX signal (the reverse of what we do in the TGS detectors).


-SB

hi.
thanks simonbaker "direct hit !!"

I want to tell same of bottom drawing.

What is the advantage- disadvantage of this idea?

Mikebg thanks for reply and effort I am read your all messages and follow you in forum
you continuously write here "twin loop" "balanced coil" please post here one schematic in the your all idea but complate .I am build it and I say it MTLDD means "Mikebg 's twin loop deep detector"

best regards.

[mikebg]

Wolf, your idea has all drawbacks described for the upper block diagram in posting #2 .
Here is a TWIN LOOP design, which operates with pulse induction
http://www.geotech1.com/forums/showthread.php?t=13622

[simonbaker]

hi.
thanks simonbaker "direct hit !!"

I want to tell same of bottom drawing.

What is the advantage- disadvantage of this idea?

Mikebg thanks for reply and effort I am read your all messages and follow you in forum
you continuously write here "twin loop" "balanced coil" please post here one schematic in the your all idea but complate .I am build it and I say it MTLDD means "Mikebg 's twin loop deep detector"

best regards.

I'm not sure how to analyze the comparison. But here are some thoughts:

With a true "lock-in amplifier", there is no "sync pulse" -- the two signals are just multiplied together (Tx signal multiplied times Rx signal) -- so there is no "asymmetry" to even talk about.

With our practical synchronous detectors, we turn one of the signals into a binary "sync pulse" and multiply it times the other signal. Your question is: which signal is best to turn into a sync pulse, the Tx or Rx signal?

It would be interesting to analyze it mathematically to try to show the difference. I suspect it is possible, but I'm not able to myself at this time!

To get started, we can think of the Tx signal as a pure sine wave. The Rx signal is a combination of several signals:

1. The "null" signal, which is a pure sine wave with same frequency but usually different phase and smaller amplitude compared to the Tx signal.

2. The "target" signal which is a pure sine wave with same frequency but usually different phase and smaller amplitude compared to the Tx signal, and often much smaller than the null signal.

3. Noise signal. For a typical MD such as the TGSL, the noise is fairly narrow band, possibly centered on the Rx coil resonant frequency.

If we convert the Rx signal to a sync pulse signal, the pulses will reflect the characteristics of the Rx signal components, for example jitter, varying pulse width, phase and frequency characteristics, stochastic variations.

However, when we multiply the Tx signal times the Rx sync pulse train we still may get a meaningful detection signal. I suspect it would work to some extent. (I think maybe mikebg is thinking of extracting a stable signal from the Rx signal to use as a sync pulse, but in this question we don't want to do that, the variations are what we are trying to detect).

For example, if both the Rx and Tx signal were square wave signals with no noise, then I don't think it would matter which one you used as the sync pulse in our Synchronous Detector. Changes in the phase of the Rx signal would be detected the same either way.

If the Rx signal contained noise that was broadband (high frequency) and fairly large, then it could lead to an impractical "choppy" sync pulse signal.

By using the Tx signal to generate the sync pulse, it is easier to analyze, and we can understand how noise frequencies away from the Tx frequency are effectively averaged out.

My intuition is that you lose information when you use the RXxsignal as a sync pulse. The Tx signal, being a single frequency, really has the same information whether you turn it into a sync pulse or not. But the Rx signal will lose information if you turn it into a binary signal -- I think that may be the compelling reason to use the Tx signal as the sync pulse.

But I can't say that using the RX signal for a sync pulse wouldn't work. You could give it a try and let us know!

-SB

[mikebg]

Here is the patent mentioned in posting #2:

[WM6]

Here is the patent mentioned in posting #2:

US4,881,031 interesting too

[lonelyWOLF]

hi..

I began to realize my idea on the breadboard.

I wonder what the result will be..?

[lonelyWOLF]

hello forum..


yes it worked .
it has very noisy signal.

finally..
result is not bad I use unreg. power supply,no shield ,ie, roughly.
upper photo
no target

bottom photo
20mm coin at 25cm distance

regards...Dagci.erol

[simonbaker]

hello forum..


yes it worked .
it has very noisy signal.

finally..
result is not bad I use unreg. power supply,no shield ,ie, roughly.
upper photo
no target

bottom photo
20mm coin at 25cm distance

regards...Dagci.erol

Well, you proved something. I don't recognize those signals, you would need to show the circuit you used and where you made the measurements.

Thanks for interesting experiment!

-SB

[lonelyWOLF]

hi

simonbaker here is my breadboard test circuits ..