Oscillator power!!!!

[Geo]

I must Thank all here for your answers.
Maybe the rectifier bridge inside the coil don't allow me to increase the signal more than 50% up. With this signal i did not see any change in the depth with small or big objects. So i will close the box of my Sovereign, But i will watch your so interesting discussion

My Regards

[Geo]

Hi Geo. "Anker SS60" has a good detection of the air test, but that proved to be the ground? Is it stable? At what depth in the soil to detect coin 25mm, larger items ...?

Hi.
I can't tell you exactly about the depth in the soil because i did not worked it for long time. With simple words i will say that it works near the "limit"... better depth than the other detectors, not so good GB and sensitive to rocks.

[mikebg]

I must Thank all here for your answers.
Maybe the rectifier bridge inside the coil don't allow me to increase the signal more than 50% up. With this signal i did not see any change in the depth with small or big objects. So i will close the box of my Sovereign, But i will watch your so interesting discussion

My Regards

Geo, since the RX preamp is placed inside sensing head, you can measure received signal at connector and check for linearity. If there is no saturation, increasing supply voltage twice shoud lead to twice increasing of preamp output. Can you measure at connector the voltage across TX winding?

[Geo]

Geo, since the RX preamp is placed inside sensing head, you can measure received signal at connector and check for linearity. If there is no saturation, increasing supply voltage twice shoud lead to twice increasing of preamp output. Can you measure at connector the voltage across TX winding?

Hi Mike.
I measure the voltage across the Tx winding at the connector with oscilloscope.
So, when i increase the power supply to the out stage from 10V to 15V i see the signal to increase 50%. After this voltage the signal remain the same even if i will supply with 20 or 25V.

Regards

[mikebg]

If this is the circuitry inside sensing head, it can not limit amplitude.
Perhaps TX circuit makes limiting of output.

[mikebg]

I found circuit diagram containing and TX end stage. It also can not limit signal because operates as bridge switch.
http://www.geotech1.com/forums/showt...eferrerid=2910

[maikl]

Hi.
I can't tell you exactly about the depth in the soil because i did not worked it for long time. With simple words i will say that it works near the "limit"... better depth than the other detectors, not so good GB and sensitive to rocks.

Thanks for your reply Geo, I come to the conclusion that the most "sensitive" detectors have a problem with stability, discrimination. Stability on the ground is paramount.

[Geo]

If this is the circuitry inside sensing head, it can not limit amplitude.
Perhaps TX circuit makes limiting of output.

Hi Mike.
Yes, this was the circuit inside the original 10" coil.
Tx end stage is the same with this that ivconic attached here. I supply the Tx end stage from one seperate power supply, so i can give it what voltage i like.
Now i use a WOT Platypous coil, so i don't know what circuit it using inside the coil.

Regards

[Geo]

Thanks for your reply Geo, I come to the conclusion that the most "sensitive" detectors have a problem with stability, discrimination. Stability on the ground is paramount.

Hi.
I see logic the more sensitive ... more problems.
If Anker had a external and different GEB then maybe to had less problems, or maybe to was the best detector.

Regards

[Carl-NC]

The 64 fold applies only in the infinite distance to the coil (far distance region). In the near detection distance region (we usually have), the factor is less, which makes sense to increase the TX power.

I'm not sure if your use of "near region" and "far region" are meant to be the same as classical electromagnetics, but metal detectors always operate in the near-field region where the signal strength of an isotropic radiator falls off as 1/x^3. By comparison, classical radiowaves operate in the far-field region which is 1/x^2.

The thing to note is that detector coils are not isotropic, nor is the target eddy response, so I would expect the round-trip signal will never be 1/x^6. I would also expect it to be non-linear for reasons you pointed out.

- Carl

[mikebg]

REGIONS at metal detecting

Current loops used for metal detecting not operate as antennas, ie in practice both the TX coil and eddy currents in target
not radiate EM wave. To calculate axial induction 'Ba' caused by a current 'i' in a point 'P' placed at distance 'z' from current loop having radius 'R', we use the formula given below.
Near region of a current loop means n<1 or point P is in distance z less than loop radius R as shown in attached
image. Because of that we can eliminate z from denominator and obtain a simplified formula (see below) .
Far region of a current loop means n>1 or point P is in distance z greater than loop radius R. Then we can eliminate
R from denominator and obtain other simplified formula (named 64th law of Murphy :-).
Experiment described in posting #15 shows that small targets (even shallow buried) operates in far region. According
to Murphy's law for the world vileness, this can not be avoided in receiving path of energy (target - RX coil).
Remains to make large TX coil to excite targets in near region. However at large diameter of TX loop (according to
world vileness :-), conductive soil appears in near region and will generate larger GND signal having timeconstants near to targets timeconstant.
CONCLUSIONS:
Every metal detector needs a family of TX coils having different diameters.

[Qiaozhi]

Current loops used for metal detecting not operate as antennas, ie in practice both the TX coil and eddy currents in target
not radiate EM wave. To calculate axial induction 'Ba' caused by a current 'i' in a point 'P' placed at distance 'z' from current loop having radius 'R', we use the formula given below.
Near region of a current loop means n<1 or point P is in distance z less than loop radius R as shown in attached
image. Because of that we can eliminate z from denominator and obtain a simplified formula (see below) .
Far region of a current loop means n>1 or point P is in distance z greater than loop radius R. Then we can eliminate
R from denominator and obtain other simplified formula (named 64th law of Murphy :-).
Experiment described in posting #15 shows that small targets (even shallow buried) operates in far region. According
to Murphy's law for the world vileness, this can not be avoided in receiving path of energy (target - RX coil).
Remains to make large TX coil to excite targets in near region. However at large diameter of TX loop (according to
world vileness :-), conductive soil appears in near region and will generate larger GND signal having timeconstants near to targets timeconstant.
CONCLUSIONS:
Every metal detector needs a family of TX coils having different diameters.

The formula you have shown is the Biot-Savart Law equation, and you have correctly stated that a metal detector coil does not operate as an antenna. Although you may have noticed that several metal detector manufacturers erroneously still refer to the coil as an antenna in their literature.

I think you should change your terminology of "near field" and "far field", as this conflicts with the classical electromagnetics definition (previously mentioned by Carl). Clearly you are really referring to the two z-axis regions either side of a point R, which is equal to the coil radius.

Anyway, regardless of the precision of the "6th power" law, it is evident that even large increases in TX power (such as doubling the output) will not provide a significant increase in depth.

[Skippy]

Hello Geo / everyone. I am experimenting with depth improvement of my own detector, and have examined the published articles (particularly Bruce Candy's), and performed reasonably accurate experiments too. I wonder if Bruce's formula may contain an error. The formula he gives for the field seen by the coin seems correct. However the return signal generated by the coin should vary with the COIN radius, call this A, and not the coil radius,a. Hence the total signal return is the product of the two expressions. For a typical detection limit of 300mm away from a 200mm coil, with a 25mm coin, it can be seen that the reciever is in the far-field of the coin (300mm >> 25mm) and hence the signal can be assumed to fall away at an inverse cube rate. The transmitted field seen by the the coin is in the 'intermediate zone', not the inverse cube of the far-field, but not the inverse square of the near-field. I would approximate this as 'inverse 2 1/2'. hence for the typical coin-sized object, I would approximate the total signal fall-off as an 'inverse 5 1/2 ' power relationship. For very small objects, eg. nuggets, ear-rings, that can only be detected 100mm from a 200mm coil, I would use the approximation 'inverse 5th' power. My experiments have consisted of varying the front-end gain of my VLF/IB detector (easier than varying TX signal) and measuring the detection limit for various size/type of objects.My coil is 200mm. I found the above quoted relationships seem correct. Tiny objects definitely benefitted more than larger objects. Example: with a gain increase of X3: large copper coin( 28mm) gave a distance increase from 260mm to 315mm,(X1.21), a small brass nut improved from 118mm to 153mm (X1.29). The approximate theoretical distance increase figures I calculated were X1.22 and X1.25 for these two cases.However, with reference to Geo's question, I would say increasing TX level would be better than increasing RX gain, due to gains in signal/noise ratio, and no extra pickup of interference, other nearby detectors signals etc,etc.Hope this enlightens you.

[mikebg]

The formula you have shown is the Biot-Savart Law equation, and you have correctly stated that a metal detector coil does not operate as an antenna. Although you may have noticed that several metal detector manufacturers erroneously still refer to the coil as an antenna in their literature.

I think you should change your terminology of "near field" and "far field", as this conflicts with the classical electromagnetics definition (previously mentioned by Carl). Clearly you are really referring to the two z-axis regions either side of a point R, which is equal to the coil radius.

Anyway, regardless of the precision of the "6th power" law, it is evident that even large increases in TX power (such as doubling the output) will not provide a significant increase in depth.

Thanks Qiaozhi, you are correct.

The attached image illustrates principle for measurement of TX path (TX coil - target).
The measurement of RX path (target - RX coil) is ilustrated in posting #15.
According 64th law (of Murphy :-), doubling distance z in TX or RX path should decrease signal 8 times or 18dB. Since energy traverses both paths from TX coil to RX coil, the doubling of distance should decrease signal 8x8=64 times or 18dB+18dB=36dB, however for precise measurement, you should make smaller increase of z, for example 20% (1.2 times) instead 2 times. Await as result signal decreasing 73% (1.73 times) or 4.75dB in RX path.
You can make measurements using computer as function generator and oscope.
Radio amateurs use for similar measurements this software:

1. Soundcard Scope of Christian Zeitnitz:
http://www.zeitnitz.de/Christian/scope_en

2. SpectrumLab of DL4YHF,Wolfgang Büscher (Buescher):
http://www.qsl.net/dl4yhf/spectra1.html#download

3. TCUBE and other instruments developed by Alberto, I2PHD:
http://www.weaksignals.com/

[moodz]

A proper H field "coin sized" sampling loop should be constructed with appropriate faraday shielding. Unless you have a differential input on your scope ( not many people have those ) the coil will pick up E fields from the TX coil AND/OR its sheilding OR surrounding environment. Should also put a differential buffer amp at input of single ended scope input.
Moodz.