ERIC FOSTER Please Respond.

[Anonymous]

I had an idea last week, and have been playing with it in the lab. Not really sure of what I'm seeing, but the results are interesting.
I have a quad transistor bridge type output stage, using a +ve and -ve supply of 9.6V from 16 C cells.
I drive two transistors to saturation, i.e. one at the top on the "left", and its opposite on the
bottom on the "right", see what I mean?
The coil is mounted BETWEEN the two halves of the output stage, and a differential anplifier used to take off the signal.
After seeing the usual exponential decay following the positive going transient, I then drive the OPPOSITE two transistors with a pulse of opposite polarity, thus giving an EQUAL, but OPPOSITE flyback pulse, and hence decay.
I have seen that, according to the "conductivity", and this is the ONLY thing I can attribute this to, without getting involved in complex hysteresis calculations, that the SECOND decay, is different depending of the nature of the target material.
I think it may have something to do with the "speed" or, as someone put it, "magnetic fluidity" of the target.
What I think I am really seeing, is the ability of the target to have it's magnetic field changed rapidly, the more ferrous the target, the less rapid the change and thus smaller the response.
Comparing the two responses, via fourier analysis (even just integrating, provides a positive signal, and a negative signal which when added, or susbtracted, take your pick, give a dicriminated signal, the larger the signal, then more conductive the item) might be interesting.
Perhaps you could try this for yourself, and confirm what I have seen.
I find that the delay of the second pulse from the first, is the MOST significant factor involved, Goldquest timings were used as the basis, with the second "anti" pulse being triggered between 26-50% of the way through the normal cycle "dead band".
I also found that this system did not suffer from saturation of the ground mineral matrix, even when I upped the TX current to some 4 amps.
I am going to withold details of the "in air" responses I was getting, as I want someone else to say what they got. Last time I posted some results on another forum, I got flamed as no-one believed them.

 

[Anonymous]

Sorry, Sean, I don't have time to build a test setup. Maybe someone else will, and report their results here.
I believe that the following basically confirms what you have already observed. Perhaps it will clarify it for other readers of this forum. If I blew it please post a correction.
As you increase the asymmetry, moving the "anti" pulse closer to the "normal" pulse, the response from low conductivity targets to the two pulses will remain about the same except for opposite polarity, of course.
In the case of high conductivity targets and iron, the response to the "anti-pulse" will weaken dramatically as you move it closer toward the "normal pulse", whereas you will see a slight increase in response from the "normal pulse".
By demodulating the short delays full wave, you can cancel earth field for the low conductivity targets. To cancel earth field for the high conductivity targets you can subtract the signal just prior to the "normal" pulse from the signal after the "normal" pulse, if you are doing analysis in the time domain (which it sounds like maybe you aren't). A discrimination signal can be produced by summing the short delays after the first and second pulses in a desired proportion, and summing/subtracting the late delay in the proper proportion and polarity in order to cancel earth field.
I'm curious, if you're willing to answer the question.. just what did you say that got you flamed in another forum? (I'm not interested in what the flamers said.)
--Dave J.

 

[Anonymous]

Hi Sean,
Sorry about the delay in replying, just very busy at the moment. There are all sorts of alternative methods of pulsing and this is one that I have not tried. i.e. moving an opposite polarity pulse in relation to the first. I know that a quad bridge has been tried by another contributor to the forum, but not the other technique, as far as I know. Unfortunately I am fully committed for the next three weeks and doubt that I will have any time to try it. However, I am keen to try as soon as I can, although I expect someone else will have a go in the meantime.
Regarding your

 

[Anonymous]

Hi again all,
My turn to apologise for the delay, but my computer crashed BIG STYLE, and I ended up losing ALL my test results.
Yes, Dave, you are correct, in your interpretation, I found that ferrous items produce a lower effective response on the "anti pulse" portion. I believe the correct term is due to the "reluctance" of the material.
Looking at the BH hysteresis curves of various materials will explain why this condition occurs.
As iron is easily magnetised, reverse pulsing it rapidly afterwards (you might like to make the SECOND pulse a lower amplitude - I havn't tried this yet) attampts to null out the magnetic field of the target and coincidentally the eddy currents within it. As ferrous materials are inherently resistive in nature, and also have a long magnetic time constant, the second pulse cannot cancel out the initial pulse completely, and so gives a residual small, but measurable difference between the two.
The technique I used was to use a low droop "sample and hold" instead of sampling FET to capture the "positive" response, and the same for the negative response. Feeding these two residuals into a summing amplifier gives the "difference" signal.
The only problem I encountered was that the result gave a negative indication, in that the higher the voltage output, the more ferrous the target. Solving this would be easy, but I ran out of time.
As regards the other forum, for a start a discriminating PI was pooh-poohed, as many stated that if it could be done, Mr. Foster would have cracked it by now. It's amazing, just how strongly people will defend Eric's reputation (and rightly so).
The other reason was the "depth for amps" results.
I found that an added offshoot of using this system, was that the "double punch" seemes to give better sensitivity at depth. This may be an anomaly in my results but means that you can use a far lower TX pulse current to get the same depths- something you might like to confirm also.
Eric, perhaps it should be described as "Almost the ultimate". I can't describe here why, but i have included one of my email addresses so that you may contact me if you wish.

 

[Anonymous]

Me again,
dave, i hadn't got round to the earth field bit, but your explanation sound VERY valid.
That problem would have been next on the list of "to do". But you seem to have saved me the effort. CHEERS <IMG SRC="/forums/images/wink.gif" BORDER=0 ALT="">
Combining both systems could make a "goer", just wish I had more time to try it.

 

[Anonymous]

Sean:
Well, if the flamers were saying discrimination is categorically difficult or impossible, all I can say is that anyone who spends time designing PI's or even reading the literature, or even bothers to read the advertisements for PI's actually being marketed, figures out that PI's can discriminate.
The events of Sept. 11 give one pause to reflect on human nature. I've come to the conclusion that flamers, all the way from beeper forums to international terrorists, do as they do because they cling passionately to beliefs which they already know are false or unsubstantiated. Your example of the forum flamers seems to confirm my theory.
Meanwhile, from philosophy back to things that go beep........ The difficulty is in getting good ferrous-nonferrous discrimination. There are a variety of schemes which can do it, but most don't work very well. Some of them have been discussed on this forum. VLF/MF machines still excel in this arena.
Regarding the relationship between your second-pulse scheme and sensitivity-per-amp: without knowing what was being compared with what, I would be unable to comment, other than to say that bipolar pulsing, in and of itself, bears little relation to sensitivity. However, in practical systems, the question of unipolar or bipolar pulsing is closely tied to matters such as transmit timing, earth field cancellation schemes, and power supply and transmit driver architecture. These matters are quite relevant to sensitivity and power efficiency.
--Dave J.

 

[Anonymous]

Dave and all,
I think I've just shot myself in the foot!
I've been looking at the way the DVT technology works on the Minelab machines, and I think this is it!!!
If this is the case, then OK, I arrived at the same conclusion, but I came up with my idea BEFORE I started looking at the Minelab waveforms.
Dang it, pipped at the post again <IMG SRC="/forums/images/frown.gif" BORDER=0 ALT="">

 

[Anonymous]

It wasn't the last good idea in the world, Sean.
May you quickly discover more.
--Dave J.

 

[Anonymous]

Cheers Dave!
I'm heartened by this attitude.
These posting are geting out of order, but I wanted to put this in here.
Yes I agree some of what I said might put people off, but then I felt I was making a statement of intent.
I have no "pedigree" such as a respected and experienced person like yourself. I have to try to make a name for myself the hard way. I now realise that, with hindsight, some of my postings have been rash and immature to say the least.
I believe VERY strongly in my abilities though, and feel that I have it within me to bring to market what I feel is an innovative and very effective design.
I value input and comments from people such as yourself, as I respect your opinions.
There would appear to be a "devil's advocate" at work in one of my other postings, and this is exactly the sort of negative comments I've been citing.
I'll keep experimenting wiht ideas, but to be honest, I'm more of a practical than theoretical engineer. I find something that works, but invariably can't explain why!

 

[Anonymous]

Sean,
Work on the leading edge is always full of puzzling observations and surprises. And, the leading edge itself doesn't really much care who has "pedigrees" (which, by the way, I didn't know I had), and who doesn't.
If what you have is in fact an improvement on what's presently being done, I hope it finds its way into the marketplace one way or another.
--Dave J.