Looking to better understand what detectors do with signals, once received...

After answering a post here about "gain" and "sensitivity," it got me to thinking...I would really like to know more clearly what happens to a target "signal," once the machine receives it. I am talking about the "target signal" only, after the "ground" signal has been "removed." I am certain that exactly what happens to the signal upon reception is different, depending upon the specific machine/manufacturer, but if anyone can help here, that would be great. I am interested most specifically in an FBS unit, but if you discuss a different unit, and specify which machine is being discussed, I'd still be interested.

I'm not looking for any highly-technical or proprietary info (like, the intricacies of how a particular "discrimination circuit" works), but simply what happens, in what order, to a signal -- from the time the receive coil "receives" it, to the time it is audibly (and digitally, in this case) presented to the user. Any type of "step-by-step" description would be much appreciated! Obviously, different things happen depending upon mode (all-metal, disc. mode, etc.); I'm simply looking to understand the basics a little better. Knowing what is happening, and how gain, senstivity, increased or decreased discrimination, etc. etc. will "change" the signal, and affect what is presented as "output" to the "user," would be helpful to me, I think, in terms of understanding what my machine is doing, based on what signal has been received. I could see potential avenues to improve my detecting if I understood what's going on "inside the box" a little better...

Thanks,

Steve



Edited 1 time(s). Last edit at 02/20/2012 03:44PM by steveg.

It's very difficult to explain in a non-technical way, and many non-technical explanations I have read are horribly wrong and misleading.
Different manufacturers/machines work differently, analogue does things differently to digital. For Minelab-specific (and general) info, take a look on Minelabs' website, there's a couple of .pdf's by Bruce Candy that are quite informative.
Carl Morelands' Geotech site has most of the patents related to MD's, but they are heavy on the technical and cover-all-eventualities legal speak.
[www.geotech1.com]
Re. your quote "ground signal has been removed", this doesn't happen. Ground signal is always there in huge amounts, clever tricks reduce/control its level, but its presence is what limits ultimate target depth.



Edited 1 time(s). Last edit at 02/20/2012 05:23PM by Pimento.

Pimento --

I do understand I'm asking a tough question...asking to put this into "non-technical" terms. I appreciate your point on that.

I'll check out Bruce Candy's stuff; in the mean time -- what I meant by "ground signal has been removed," wasn't quite the way it came out. Maybe a better way for me to have said it is "the part of the signal that the machine has separated from what it thinks is the ground return." I realize it's not "removed," per se; my understanding is that the machines try to "pick out" any possible "non-ground" signal from within the much stronger "ground return." Is this a correct understanding? What I was simply trying to say was that I didn't need anyone to go into the "hows" of ground balance and the ground signal; I think I have a general idea of how that all works; I was just trying to say that I was wanting to "follow the process," so to speak, from a time point AFTER the machine thinks it has received a "non-ground" (metal object) signal...

Thanks!

Steve

there is a good explaination here. [www.nexusdetectors.com]

Tom in SC

Steve, FBS uses a unique way of measuring the the FERROUS properties of a target. On the last page of MineLabs' document at [www.minelab.com] they say measurments at the end of the received wide pulse (as shown in the figure) will produce "excellent information about the ferrous nature of metal targets". My understanding is the deeper the slope of the received wide pulse, the higher the ferrous properties of the target. In the ETRAC, after the end of the wide pulse is measured the computer assigns an FE value between 0 and 35.

I am still trying to figure out how they use the rest of their transmitted signal to sort out the CONDUCTIVE properties (CO) of a target.

Tom --

Thanks much. That was a very good article, not too technical and a great read. Enjoyed it.

I wonder, though, what happens AFTER that point...once the receive coil receives the signal...

Another way I could clarify what I'm trying to figure out is -- I have heard, as just one example, the term "pre-amp gain;" what does that mean, in terms of the signal, as the signal is being processed? How does it differ from other types of "gain?" Along those lines, what about "gain" and "sensitivity?" I understand these two terms a little better, but still -- it would be interesting to understand all of these things a little better. Additonally, things such as "discrimination circuits" and "filters" and such; I hear of terms such as "two-filter" and "four-filter" but don't know what that means. I realize that discrimination will affect what is allowed to be "passed through" to the user, but is that a "filter?" Basically, I know a little about some of these terms (again, such as "gain" and "sensitivity" as they relate to an Explorer); other terms, though, I know very LITTLE about (again -- like a "filter" -- what is happening to the signal as it goes through a "filter?") Etc. etc. etc.

Basically, to create a "word picture" of what I'm imagining, I'd like to, in my mind, be able to "ride along" with a signal, as it enters the receive coil and then moves through the electronics of the machine -- such that I can understand what each "step along the way" does to the signal -- up to the time it is "output" to the user as a number or a tone (or both).

Yeasty -- thanks for FBS/FE link; I'll "dig in."

Steve

Yeasty --

I'll need to re-read that Bruce Candy article several times -- but it was very, very interesting. I gained alot from it...

One thing I found interesting was that for a single-frequency VLF detector, he said that changing the ground balance has no effect on ability to discriminate targets (instead -- that ground balance only affects a detectors ability to separate the "R" component of a target from the "R" component of the ground, when the target is deep, or small, when running in all-metal mode). That's how I understood what he was saying, anyway...

Fascinating...

Steve

Most technical terms related to MD's that you have read about are probably not applicable to the fbs machines. They are 'dual-frequency', for a start, so how they distinguish a target from the background is very different to a single-freq machine.
It's also very easy for a manufacturer to label some feature with an inaccurate but user-friendly name. A knob labelled 'Power' invariably does nothing of the sort. I suspect Ferrous and Conductive fall into that category. If they called it 'Normalised ground-compensated low-frequency demodulated phase' or some-such, you would be baffled! Throw in some digital/microprocessor/DSP stuff and even something like 'Gain' might not change the gain of anything.
Regarding a few of your specific queries:
Sensitivity: discriminating machines generally look at the heavily-processed receive signal, and see if it exceeds some preset threshold level. If it does, the machine will beep at you. This threshold level can be varied. Set it high, and a bigger target signal is needed to get a beep. Set it low, a smaller target signal will give the beep, along with random ground signal variations, nearby cellphone antennae, etc.
Differentiator - a type of filter: Your machine sees the constant fluctuations of the ground mineralisation/iron contamination. When you pass the search-coil over a target, the signal increases rapidly, then drops back to the background level. A differentiator is a doohicky that responds to rate-of-change of something. It accentuates changes, whilst not responding much to the constant background. In an MD, it emphasises the signal increase/decrease caused by the target passing under the coil. This makes it easier to tell if there's something there or not.
Gain: could be just about anything, depending on what signal you are changing the gain of (if at all)
Two/four filter: filters let certain signals through, blocking others. Careful chosen filters will allow 'good' signals to be seperated from 'bad'. The nature of the good and bad determines which type of filter is best, a modern machine will choose on its own usually, so you don't need to care or know.

Thanks for the additional info, Pimento. As you said earlier, hard to take the complex question I asked and put it into a nice, simple package...

One thing I didn't know -- you say that sensitivity determines the level of the POST-PROCESSED signal, and not the PRE-PROCESSED signal...didn't know that. Interesting...I thought it was the other way around...

Steve

Trouble is, it could be 'the other way around' to a certain extent. If you boost all the signals (front-end amplification) you will probably make the signals later-on bigger, too. But not always, due to compression/limiter circuits. You could do both, of course. The Fisher F75 changes front-end gain when you go from the '1 - 29' sensitivity to the '30 to 99' sensitivity. The manual mentions this, in fact. This would normally not concern the user, but I can confirm that in extreme iron laden ground, the detector 'overload' warning comes on less easily when sens. is below 30.

Very interesting...I can see I opened up a real "can of worms" here...that has no simple explanation.

I guess I feel like it would be helpful to understand what is going on...I remember starting out how confused I was by the terms "gain," "sensitivity," "threshold," etc. etc., and while I understand things quite a bit better now, I really would like to understand even more clearly how things work...I really feel like understanding the IMPLICATIONS of what's going on, could really help someone use their machine most effectively.

Steve