Latest Minelab Patent - 11th Feb 2016

[Carl-NC]

I never met Don, but I did get the chance to dig through his notebooks. I liked his work. I was told the MF design he did worked brilliantly when it worked, but it didn't work often enough and the project dragged on too long. (Unfortunately the prototype had disappeared.) It was, in my opinion, too far ahead of the chips that were available at the time. With DDS and powerful 32-bit micros, it might be the kind of project you'd resurrect and turn into a commercial product.

[Carl-NC]

Except for occasional effusion of unnecessary arrogance by some people, I do not see anything special educational in this whole debate.

The reason I keep coming back around to the BBS transmitter and how it works (as opposed to how people believe it works) is because, for those who are interested in moving to a MF design, it's a really good way of getting there. With just a micro and a few MOSFETs, you can have a very flexible transmitter that can do 1, 2, 3, or more frequencies, in any ratio you like, all selectable at run-time. Add some vanilla demods and the rest is Fun with Signal Processing.

But here's another tidbit: The Impulse transmitter design can do what the BBS transmitter can do, but more. This becomes apparent by playing around with it, preferably in Spice as Davor suggests. It can even produce a simultaneous hybrid VLF+PI waveform. With a little cleverneering, it will do multi-frequency hybrid VLF+PI. The prototype I had running was pretty impressive and exciting, and one day I may resurrect this concept as well.

[Beav]

Carl. I got a Kick out of your term "cleverneering"! It suits engineering to a "T"! LOL
Thanks for making me chuckle!
Beav

[kt315]

half brigde + half bridge you get a bridge

this driver from Koschej. you can use it with resonance coil or un-resonance coil. you may call your detector either PI or hybrid - on your choice.

[clancy]

Teleno, you do what many do on forums. Spot few “key-sounding” sentences and post these as “proof” paper or patent say that you want forum viewers to believe it say. Lucky in academic scientific nobody can cheat like this.
You say “TC spectrum” taught in this paper “UXO time-constant estimation from helicopter-borne TEM data” on page 8, and Wei, Scott, McClellan paper “discloses ground balance using the same technique.” This be not true. What Wei paper do, is see if Rx signal = soil well-known log-uniform VRM signal and if it do, then no target present in Rx. Only if Rx be different to soil log-uniform VRM signature, only then must be target detect, and only then data use for generate target TC spectrum from Rx data without any ground balance.

Thank you for last bonus paper. Note this written by same author Wei of paper above. It also do not ground balance and paper say problem of exactly what the patent solve! It say “The response due to the magnetic properties of the soil can also influence the DSRF since the response of the soil is superimposed on the response of the target and it is not possible to completely separate the two.”
Excellent proof that how to GB and then generate TC spectrum be NOT OBVIOUS. Thank you Teleno for this proof. If you still think GB then transform to TC spectrum be obvious, maybe you write to Wei, Scott, McClellan and tell them they stupid?

I see you cannot answer my questions to you. Only if you can, only then you understand why patent solve problem of ground balance then transform to TC spectrum for first time be not obvious. The 2 numbers ask you be very different to each other.

If you read patent you see it point out problem that Wei (et al) (who you cite twice) cannot solve:
“It is known to demodulate receive signals of a metal detector for processing to characterise the receive signals in terms of time constants. WO/2011/011820 discloses that discrimination of a target to a certain degree can be achieved by analysing the spread of the time constants. Mu-Hsin Wei, Scott, W. R., McClellan, J. H., “Estimation and application of discrete spectrum of relaxations for electromagnetic induction responses,” Geoscience and Remote Sensing Symposium, 2009 IEEE International, IGARSS 2009, vol. 2, no pp. II-105, II-108, 12-17 Jul. 2009 discloses that by identifying each of the main frequency components (inversely proportional to time constant) of the received signals, it is possible to enhance discrimination. Mu-Hsin Wei, Scott, W. R., McClellan, J. H., “Landmine detection using the discrete spectrum of relaxation frequencies,” Geoscience and Remote Sensing Symposium (IGARSS), 2011 IEEE International, vol., no., pp. 834,837, 24-29 Jul. 2011 discloses a pre-screener to ignore cases where substantial soil signals are present.”

Also, “UXO time-constant estimation from helicopter-borne TEM data” have no ground balance, and do not generate TC spectrum only single TC in results (but do mention TC spectrum), and it not sync demodulate then LPF, but get TC from time data, but it say that this technique have poor s/n: This be why patent synch demodulate then LPF to get good s/n.

[daverave]

Except for occasional effusion of unnecessary arrogance by some people, I do not see anything special educational in this whole debate.
I think that it is precisely the presence of this arrogance prevents that the debate goes in the right direction.
Egocentricity and arrogance otherwise spoil the many topics on this forum.
So that all of this became very boring.
It all boils down to masturbation some of these arogant people who seem to have no other ways of pleasure in life.
Whichever professional achievements that man has; become completely irrelevant and void by the presence of such arrogance.
And such people are good for avoiding and ignoring.
The theme revolves around, nothing new has been said, nothing was found.
People from ML are probably lauhing to death reading this.
So this is all very comical and pathetic.
And those who were arrogant, and who think of themselves are worth more than others; better yourself sit down and learn something more than what you know.
It's one thing to know for real and another to be the turkey struts with colorful feathers, tossing empty words around.
When you can not find the right answers; then along comes this course of behavior and action.

Whether he was joking or not, KT315 gave here pretty sane and honest question and doubt.
In the spirit of this forum, it should be just tried to answer on that.
Several nice people tried to do that.
Some of not so nice people took advantage of the situation to show their arrogance which is also their weak point.
Now we know that from such can not expect any contribution here.
I can not participate in such a story, i'm sick of these kind of people.
And my last advice would be; do not continue to give cause ML people to laugh!
They did a job and now they taking a money for that... while you here are wasting your (and our) time by empty debating.



https://www.youtube.com/watch?v=XKk5KGp6Foo

i find some of the input of this discussion very daunting...ok there are some high tech people on here that do this for a living but it makes one feel very small in comparision when some of us are mere mortals and just hobbyists....even with all these big statements of high tech jargon what does it all add up too.......how far have modern motion machines and such like have come ???? take away fancy lcd displays and gimicks compared to old technology how far have we really come ???? but marketing wants to make us believe the la
test tech is the best...but in reality maybe not !!! even home hobbyists can equal and better the depth on high end machines at a fraction of the cost........i know of someone who used an old arado 95 and he has found more gold over the years with that old machine than many will find in a life time with their modern dual frequency...bbs...dvt....vflex and such like.

[Davor]

Nope. Voltage supplied to the inductive reactance is what does it. The physical coil has series resistance -- plenty of it-- and some of the voltage gets dropped across that. To make matters worse, in the case of a perfect inductor knowing the voltage at any point in time tells you nothing about the current because the two things are 90 degrees apart. If you want to know what the magnetic field is, you have to know the current. This is ordinary high school physics.

OK, I see your point. I normally don't go to the parasitics when trying to present a concept in its simplest form, but again, you are right.

My view of the practical coils is that they are mostly of considerable Q factor in a working frequency range. Series resistance is very obvious with shorted coils that refuse to produce constant field, but instead you have some reactive component.

Could you be bothered to elaborate on the making of CZ5? It is a great machine and it surely must have a story behind it. If not here on some dedicated topic?

[Teleno]

It also do not ground balance and paper say problem of exactly what the patent solve! It say “The response due to the magnetic properties of the soil can also influence the DSRF since the response of the soil is superimposed on the response of the target and it is not possible to completely separate the two.”

You keep repeating that the patent solves the GB problem but as I earlier pointed out, the patent discloses no such a thing at all.

So let me challenge you to quote the patent where it supposedly discloses GB (not simply mentions it in passing) in such a way as to allow the skilled person to duplicate the alleged technique. Once you post here the juicy bits (if you find them) we'll see whether it's known from the prior art or not.

So please, cut down on your own text and point us to the literal subject matter, if it exists.

Also, “UXO time-constant estimation from helicopter-borne TEM data” have no ground balance, and do not generate TC spectrum only single TC in results (but do mention TC spectrum)

You certainly lack understanding of what a time constant spectrum is.

The definition of discrete SPECTRUM is "decay parameters (time constants and weights)" (page 8 ). The signal is "decomposed as a weighted sum of pole responses" (i.e. a discrete SPECTRUM ):

The expression above represents a discrete time-constant spectrum. This fact flew over your head, another strong indicator that you know little about what you're talking about. Whether the word "spectrum" is mentioned or not is irrelevant, the disclosure is there and the skilled person understands.

Then on page 9:

"...two time constants were estimated ... for this example .. approximating the two strongest eigencurrent modes in the plate model..."

So you're proven wrong again, the paper does disclose a two-time-constant approximation of a target just like ML's "invention" does.

[Dave J.]

Could you be bothered to elaborate on the making of CZ5? It is a great machine and it surely must have a story behind it. If not here on some dedicated topic?

Well, this thread has already been clogged with off-topic stuff, I don't suppose a post on the CZ5 would do any additional harm. What was it you wanted to know?

[Davor]

Just the story. It most certainly is inspired, so it must have a good story.

[Dave J.]

Went to work for FRL early Feb 1981, knew nothing about VLF metal detectors. In April I wrote Jim Lewellyn (company president) a memo saying that the future was multifrequency, it would soon do to singlefrequency what VLF did to TR's and what TR's had previously done to BFO's. (With prediction skills like that, it's a good thing I don't play the stock market.) Meanwhile we needed something less ambitious, and that was the 1260-X, the first double derivative pattern "motion discriminator". After we released that in May of '82, I began playing around with multiple frequency. No Internet, nothing available to copy, I had to figure it out on my own. Took a long time, there was lots of other stuff I had to design especially for our industrial line, and we took a long detour into pulse induction (another thing I had to figure out on my own, that's why the Impulse is different from conventional PI). Some prototypes were proof-of-concept that were just for the bench, others were "swingers". The first prototype that resembled what actually went into production was HC-17, "Harmonic Classifier", my 17th multifrequency prototype.

The platform is extremely complex, half the circuitry and most of the tweaker pots could be done in software. Twice during the project I stopped to figure out what it would take to actually do that, and both times I decided that the A/D's and microprocessors then available weren't up to the task, and neither was I (never having coded a microcontroller before). It was going to have to be an all-analog machine. And, iron discrimination was going to have to be reactive, because to do it right using multiple frequency methods would have made the project vastly more complicated than it already was. We introduced it in November 1991 as the CZ-6, a waterproof package without a built-in speaker.

The quest for waterproof came to an end with the waterproof connectors, which field tested fine but didn't hold up in vigorous field use. We had to go back to the old reliable but non-waterproof Switchcraft Slimlines. And, customers wanted a built-in speaker. So first we had the non-waterproof CZ6-A with the Switchcraft retrofit, and then the CZ5 with built-in speaker that survives today as the CZ3-D, which has slightly rearranged target ID categories but is otherwise the same. In 1995 we introduced the scuba-rated CZ-20, later to become the CZ-20 when the manufacturer of the original oil-filled pinpoint switch stopped making it and we had to convert to a piezo switch.

[6666]

Thanks for the story DaveJ

[Davor]

Thank you Dave. That's inspiring.
I'm convinced your prediction of multifrequency future was correct. Apart from PI gold machines it is very true for high end machines of today. In the most parts of the world we do not have much use of gold machines anyway.

Given the benefit of time passed, what would you do differently - given the knowledge you have now?

[crane]

For those still interested in what BBS/FBS is, here is the voltage & current waveforms:



I've posted this a few times in the past but it doesn't seem to make much of a lasting impression. Look closely at the current waveform... see any 'off' time? Pretty continuous-looking, eh? This is why I posted that 2-frequency sinusoid and asked what is it? and how would you process it?

P.S. -- if you think my Visio art might not be an accurate representation, I've also measured the waveform with a current probe on an oscope. Same thing.

Carl,
Yes you have posted the BBS/FBS current waveform before but it's what we would expect, just as we wouldn't be surprised to see the SD2200 current waveform supplied by Eric in the attached picture.

Anyway, in one form of BBS/FBS BC takes a number of samples at spaced apart times. The first beginning soon after one of the abrupt transitions but after ground eddy currents have decayed (same as PI). One combination is used to cancel the historical component or VRM (same as PI) but where it differs is that he also cancels the reactive signal, which is common to all CW detectors.
Note that he mentions the same rules for the historical component as he does for PI. The VLF phase shift or lag would be explained somewhat differently.

He also points out that the ratio of the reactive (instantaneous) and historical (VRM) signals is independent of frequency up to 100kHz (same as PI) which means that RX is broadband (same as PI) whereas Dave J's device must filter the RX to admit only selected signals. He also points out that Dave's device is FD whereas BBS/FBS is TD.

You look at the BBS/FBS current waveform and see VLF, I say look at the patent and we see TD very similar to PI and nothing like VLF. The only thing certain is that it isn't traditional PI.

[kt315]

You look at the BBS/FBS current waveform and see VLF, I say look at the patent and we see TD very similar to PI and nothing like VLF. The only thing certain is that it isn't traditional PI.
---
you will second who says that. after me