Pulse Induction History and Theory

[mikebg]

=Carl-NC;164483]These early detectors (all the way back to Bell's 1881 model) were not PI but rather induction balance. They used an electromechanical clapper to generate the TX signal. If it appears that the search coils are not balanced (as on the magazine cover), there are 2 more coils in the box which are used to create the IB.

Carl,
I hope you to have enough free time because arises a long-running dispute. The reason for this is, that you and most participants of the forum are not from the amateur radio community where the terminology is established in accordance with textbooks on radio electronics. You and most experts in MD technology, use an incorrect and misleading terminology. Currently only Dave J. not lost the habit of calling things by their right names.
The bad thing is that the incorrect terminology is accepted by Qiaoshi and perceived by participants in the forum as normal. Perhaps the wrong terminology is to use in the written by both you bible.

Fortunately, the experts on metal detectors for humanitarian demining use correct terminology. As an illustration, I give an example how a ham radio designer will write the information in your posting:

The early detectors (all the way back to Bell's 1881 model) were (like a conventional PI) wideband machines. They use an electromechanical interrupter to generate wideband TX signal in audio frequency spectrum. If it appears that the search coil arrangement is not induction balanced (as on the magazine cover), there are 2 impedances in the control box which are used to create bridge balance.

(Seems as terms BBS and FBS of Minelab :-)
To make a long dispute, I suggest we to use the following thread:
http://www.geotech1.com/forums/showt...766#post113766

[Carl-NC]

What's the difference between a wide-band machine and a narrow-band machine? If I take a Bandido and remove the BPF from the RX channels is it now a wide-band machine? If I take a SurfPI and apply tight BPF to the RX is it now a narrow-band machine? The early detectors were only wide-band in the sense that they probably didn't apply BPF, but certainly could have. But they were definitely not PI, if we were to agree on what "PI" even means.

As I mentioned in another thread, terminology is often driven by marketing in the need to communicate with laymen. Unlike amateur radio, where practically every user is fairly well-versed in the technical underpinnings of the box they are playing with. If you think PI is difficult, try TR, IB, VLF, and RF. Is the Minelab Eureka or Fisher 1280x a VLF detector? Is the White's "Coinmaster TR" an IB detector? Is the SurfPI also a TR?

Good luck in getting this mess straightened out.

- Carl

[Carl-NC]

BTW Mike, you would like my current project. It is both WB and NB. It is both PI and VLF. It is TR and IB. I can guarantee the marketing dept will say it is both TD and FD, whether I agree or not. And the next-gen version with multi-frequency (or is it multi-period?) is already on the bench.

[Aziz]

TD, FD ? Ah, that's simple :
TD: Terrible Detector
FD: Fugging Detector
SD: Super Detector
*LOL*
Who really cares?
Aziz

[mikebg]

TWO OPERATING MODES

Each extremely sensitive detecting device (radar, sonar, night vision, metal detector) should operate in two steps:

http://www.geotech1.com/forums/showt...590#post101590

1. Search for target! Alarm (alert) if there is target! This step needs narrow band search procedure to obtain extremely high SNR (signal to noise ratio). The narrow band is achieved at end of 19th century with a switch S1 which connects capacitor C1 to ground. Then the ancient "function generator" delivers almost sine wave at low audio frequency.

http://www.geotech1.com/forums/showt...799#post101799

2. Idetify the target! This is wide band recognition procedure (afther target is detected and pinpointed) because the device should find difference in spectral characteristics (timbre) of targets. The wide band worsens SNR and sensitivity. It is achieved by disconnecting the capacitor C1 from ground and it appears in series with load. The low search frequency is removed and the "function generator" delivers wide band spectrum, rich of harmonics.
Note that instead cheap conventional interrupter is used expensive motor commutator M1 to achieve pulses with low duty cycle.
http://www.geotech1.com/forums/showt...008#post102008

I will illustrate in complex plane why different targets have different timbre in audio frequency spectrum.

[bbsailor]

Each extremely sensitive detecting device (radar, sonar, night vision, metal detector) should operate in two steps:

http://www.geotech1.com/forums/showt...590#post101590

1. Search for target! Alarm (alert) if there is target! This step needs narrow band search procedure to obtain extremely high SNR (signal to noise ratio). The narrow band is achieved at end of 19th century with a switch S1 which connects capacitor C1 to ground. Then the ancient "function generator" delivers almost sine wave at low audio frequency.

http://www.geotech1.com/forums/showt...799#post101799

2. Idetify the target! This is wide band recognition procedure (afther target is detected and pinpointed) because the device should find difference in spectral characteristics (timbre) of targets. The wide band worsens SNR and sensitivity. It is achieved by disconnecting the capacitor C1 from ground and it appears in series with load. The low search frequency is removed and the "function generator" delivers wide band spectrum, rich of harmonics.
Note that instead cheap conventional interrupter is used expensive motor commutator M1 to achieve pulses with low duty cycle.
http://www.geotech1.com/forums/showt...008#post102008

I will illustrate in complex plane why different targets have different timbre in audio frequency spectrum.

Mike, Eric, Aziz, Carl, Al

I see the future of metal detectors being an integration of what has been individually developed and proven but fragmented and not yet fully integrated. Why just two modes?


A new line of metal detectors could evolve being microprocessor controlled that has the following modes. I agree with Mikebg that the search mode should be the most sensitive. However maximum sensitivity can only occur if the condition of the ground is known and machine parameters are adjusted to be optimal for those ground conditions (and noise conditions).

1.0 Ground Analysis Mode - With a selected coil type (Mono, DD, Concentric, Balanced, etc) the coil is lowered to the ground and the computer analyzes for the following characteristics:
1.1 Mineralization
1.2 Magetization
1.3 Delay or offset where the ground is detected
1.4 Variation within a single sweep
1.5 Sweep speed
1.6 Optional ability to store these results in the more advanced model on removable media for later analysis or parameter sharing

2.0 Search Mode - allows the user to select which features are being used to save power.
2.1 Graphics display on or off
2.2 Back light on or off
2.3 Audio output
2.3.1 Variable frequency
2.3.2 Variable amplitude
2.3.3 Sweep speed alert
2.4 Save target profile to removable media for later analysis

3.0 Optimization Mode - Uses the results from the above features along with data tables and common target characteristics to optimize the following
3.1 Search coil selection
3.2 Pulse width
3.3 Delay
3.4 Sweep speed
3.5 TX Frequency
3.6 RX gain
3.7 RX dynamic range

4. Target Analysis Mode - Once alerted from the Search Mode that a target exists, switch to the Target Analysis Mode to sweep the target from various directions at various speeds to deduce the probable identy of the target. Once the target is dug up and physically identified, the target profile could be added to the target ID library. This mode has the following features.
4.1 Target profile library
4.2 Ground profile library
4.3 Sweep speed data tables
4.4 Coil data tables
4.5 Common noise data tables
4.6 Use-added noise profiles


With a G-Force sensor on the coil shaft the swep speed could be registered and optimized for the coil size targets sought and ground conditions. It is well known that the high variability in ground conditions and the nature of targets sought can present some conditions that can be difficult to optimly analyze except by lengthy trial and error. I envision a new series of metal detectors having various amounts of the above features with the top of the line models having removable storage media to do later analysis of the following data imprinted on graphic frames containing the removable data.
1. TX PPS rate
2. RX delay
3. Ground condition
4. Date and time
5. File name for relating field-written comments
6. Found target electrical signatures
7. Coil size and type used
8. Target and ground graph

By making this metal detector modular I see optimization occuring in orderly stages. The power supply and mother board can be integrated. Then, a separate board for the following can be integrated and evolve as technology allows. The following list is notional for example only.

1. Motherboard - Power supply
2. TX board
3. RX board
4. Microprocessor board
5. A to D converter board
6. Timing Board (with G-force sensor connection)
7. Graphics Board
8. Audio Board
9. Storage media support board for removable media
10 Internal board or chip for internal data table storage


I envision software updates being made by loading a new operating system on the removable media. Internal data tables can be user-updated when people using the same detector can communicate on forums such as this and be sharing data tables that work well in specific areas while looking for specific targets. Also, when a country changes the metallic content of coins, as is being proposed here in the U.S., new target data profiles can be easily updated.

I hope this vision gets some of the best minds looking in the same direction.

Thanks

bbsailor

[mikebg]

Joe, you are great!
In a thread entitled "PI history", where all discussion the history of engineering in the deep past, suddenly appears an old man, a former strategic planner and instead of looking back in the past, he is planning the future of metal detectors. You showed that you are much younger than us. I've always admired you!

For those who like to produce fake antiques, I suggest to make this ancient project. At the moment I can not explain how it works, because for this I need complex diagram in complex plane. Soon I will draw it.

[ApBerg]

Hello,

Nice protocol from BBsailor...also add gps coord with the 4. Date and time logged record.

[WM6]

Hello,

Nice protocol from BBsailor...also add gps coord with the 4. Date and time logged record.

Agree regarding BBsailor approach, and to ad at least "Soil Wetness" under Chapter 1.0.

[Aziz]

Hi all,

Well, I strongly propose the KISS-principle. I always loved the KISS-principle. Hey, I'm in love with KISS-principle.
Aziz

[mikebg]

Here is information for variometer K3 in schematic shown in post #37.
As you see, the term "Telephone receiver" used before 100 years means headphones.

[mikebg]

HISTORY OF WIDE BAND MD (slide #1)

The hystory of DESIGN TOOL

Frequency domain is the most powerful tool for analysis and design of all types metal detectors. For example, at PI design, the tool can analyze what is the most efficient form of TX current, how to calculate delay of different samples, how PI can eliminate the GND signal, how to identify and discriminate different targets.

In UK, in US and in SPICE program, this tool is called "AC analysis". In East (relative to UK :-) Europe, it is called spectral analysis, because AC (in French CA, in German WS) means current or voltage with mains frequency 50Hz.
My first post in this forum was a question:
http://www.geotech1.com/forums/showt...7949#post67949
However, using correct terminology, the question would be:
"What are electromagnetic spectral characteristics of different soils and different targets?"

Every ham radio designer is trained to use AC analysis. There are some excellent textbooks on AC analysis in the WEB:
http://www.personal.rdg.ac.uk/~stsgr...c_circuits.pdf
http://www.ece.msstate.edu/~donohoe/...c_analysis.pdf
http://www.ing.unitn.it/~fontana/spi...C_Analysis.pdf

THE FATHER of AC analysis

The father of impedance Z=R+jX and complex plane was born in Germany as Karl August Rudolph Steinmetz
http://www.youtube.com/watch?v=d6lYoKBnPOY

He was renamed by General Electric as Charles Proteus Steinmetz
http://www.ieeeghn.org/wiki/images/1..._steinmetz.pdf
http://zrno.fsb.hr/katedra/download/materijali/966.pdf
http://venus.ece.ndsu.nodak.edu/ece/.../steinmetz.htm
http://www.schenectadyhistory.net/wo...etz-Papers.pdf
I will attach a PDF file because I don't know its URL:

[proscan]

@Prospector_all

We live in a real world, so nothing is perfect. Each forums has its non-sense, irrelevant content.

I believe if something is in foreign language (and not readable for everyone) yet still relevant and sticks to the matters been discussed is better than good-english off-topics.

By the way, is this thread still related to PI history, or I missed something?

Keep up posting Eric!

proscan

[mikebg]

By the way, is this thread still related to PI history, or I missed something?

Keep up posting Eric!

proscan

You missed frequency domain - the most powerful tool for analysis and design of all types metal detectors. For example, at PI design, the tool can analyze what is the most efficient form of TX current, how to calculate delay of different samples, how PI can eliminate the GND signal, how to identify and discriminate different targets.
Can you find errors in these loci?

[Ferric Toes]

Back to 1884

A very good paper to read, for those who can cope with a deep mathematical treatise, is “On the Induction of Electric Currents in Cylindrical and Spherical Conductors” by Sir Horace Lamb FRS. This can be found in the Proceedings of the London Mathematical Society for 1884, page 139 – 150. Horace Lamb studied under Professor James Clerk Maxwell, and also was responsible, along with Oliver Heaviside, for discovering the skin effect in conductors.

On page 142 Lamb states “There is one case, however, a very interesting one from a physical point of view, which admits of much simpler treatment, viz., that of the currents induced by the cessation of a previously constant field”. For the sake of the maths he specifies the time of the break to be infinitely short. The maths then commences.

Comments within the developing mathematics are:-

“Immediately after the break there are no currents in the interior of the cylinder (or sphere), but only at the surface, where we have a current sheet”.

“The case where the intensity of the field suddenly rises from zero to I, being constant both before and after is solved in exactly the same way”.

“When the substance of the sphere is susceptible of magnetization the work is more complicated”.

“A spherical portion of matter susceptible of magnetisation by induction is slowly brought up from a distance into a magnetic field of uniform intensity I, and then suddenly removed. During the approach of the sphere there is a gain of work………During the instantaneous withdrawal a current sheet is developed in the surface of the sphere, of strength such as to maintain the magnetic force in the interior unaltered”.

“Conversely, if the sphere be first suddenly brought into the field and afterward slowly withdrawn, there is spent in the first instance an amount of work….”

“As the currents in the sphere decay, this energy takes the form of heat”.

Note that in this paper there is no need or mention of flyback in the expressions explaining the generation of eddy currents. It is simply the sudden cessation or creation of a magnetic field and the induction effect on a conductive target.

Eric.