What are the significant differences between all metal and discrim modes on a digital machines like the T2?

[SteveP(NH)]

Besides the obvious difference of AM accepting all targets and discrim not responding to targets which are being discriminated out what is the technical difference between the two modes. In other modes what advantage to you gain from running in AM mode as opposed to discrim mode with zero discrimination set? I guess the other obvious difference is the audio is restricted to monotone, continuous processing mode but you can config the zero discrimination mode that way too.

I know in the days of analog designs that the discriminate mode was a different and longer circuit path with more components in it and so some of the signal was lost due to heat and internal noise so on an analog design the discriminate mode didn't go as deep due to this signal loss. However on a digital machine like the T2 the discrimination is done in software and so there shouldn't be any additional signal loss. The discriminate mode just runs a different program to do the signal analysis. I assume that this program would be longer due to the additional signal analysis and therefore take longer to execute but with a fast processor that amount of time doesn't seem as if it would have a significant effect on recovery time.

Can anyone (Dave J ?) say with certainty what these differences are? (that is without straying in the realm of proprietary info). BTW Dave if you see this, I hope the "Eta Kappa" project is going well )

 

[Justin E]

There is generally a significant boost in depth. I have heard of some people claiming to gain 4+ inches. I believe it also becomes more sensitive to smaller objects from what I hear. I can't say for certain, though. I haven't yet been able to afford a T2. This is based on what I've heard and my limited experience with detectors.

 

[gmanlight]

Most machines , even on zero Disc have some built in Disc .
Some more than others.
Most machines tend to go deeper in AM

 

[Elton]

Since the signal received from any given metal object exhibits its own characteristic phase shift, it is possible to classify different types of objects and distinguish between them. For example, a silver dime causes a much larger phase shift than an aluminum pull-tab does, so a metal detector can be set to sound off on a dime yet remain quiet on the pull-tab, and/or show the identification of the target on a display or meter. This process of distinguishing between metal targets is called "discrimination". The simplest form of discrimination allows a metal detector to respond with an audio output when passed over a target whose phase shift exceeds a certain (usually adjustable) amount. Unfortunately, with this type of discriminator the instrument will not respond to some coins and most jewelry if the discrimination is adjusted high enough to reject common aluminum trash for example pull-tabs and screw-caps.

A more useful scheme is what is called "Notch Discrimination". With this type of system, a notch in the discriminate response allows the metal detector to respond to targets within a certain range (such as the nickel/ring range) while still rejecting targets above that range (pull-tabs, screw-caps) as well as below it (iron, foil). The more sophisticated notch metal detectors allow for each of several ranges to be set for either accept or reject responses. White's Spectrum XLT for example, provides 191 individually programmable notches.

A metal detector may provide a numeric readout, meter indication, or other display mechanism which shows the target's likely identity. We refer to this feature as a Visual Discrimination Indicator, or V.D.I. Metal Detectors with this capability have the advantage of allowing the operator to make informed decisions about which targets they choose to dig rather than relying solely on the instruments audio discriminator to do all the work. Most, if not all, V.D.I. metal detectors are also equipped with audio discriminators.

Metal detectors can distinguish metal objects from each other based on the ratio of their inductance to their resistivity. This ratio gives rise to a predictable delay in the receive signal at a given frequency. An electronic circuit called a phase demodulator can measure this delay. In order to separate two signals, such as the ground component and the target component of the receive signal, as well as to determine the likely identity of the target, we use two such phase demodulators whose peak response is separated from each other by one fourth of the transmitter period, or ninety degrees. We call these two channels "X" and "Y". A third demodulated signal, we call "G", can be adjusted so that its response to any signal with a fixed phase relationship to the transmitter (such as the ground) can be reduced to zero regardless of the strength of the signal.

Some metal detectors use a microprocessor to monitor these three channels, determine the targets's likely identity, and assigning it a number based on the ratio of the "X" and "Y" readings, whenever the "G" reading exceeds a predetermined value. We can find this ratio with a resolution of better than 500 to 1 over the full range from ferrite to pure silver. Iron targets are orientation sensitive; therefore as the loop is moved above them, the calculated numerical value may change dramatically. A graphic display showing this numerical value on the horizontal axis and the strength of the signal on the vertical axis is extremely useful in distinguishing trash from more valuable objects.

 

[SteveP(NH)]

Hey Guys thanks for trying to answer my question - I guess I must not have asked it very well as none of the answers really answer the question I tried to ask.

Elton - I understand how discrimination works but my questions was related to the T2. What I want to know is on the T2, what is the difference between running in discrimination mode with the discrimination set at 0 so there is no discrimination taking place. So for the purposes of my question, how discrimination works doesn't really matter.

gmanlight - The T2 target ID number go down into dirt (look at the labels where the signal strength arrows show up at the top - the very bottom label says dirt and looks like 1 - 10 are dirt, not iron. So I don't think that the T2 still has some discrimination (that matters) even if it is set at 0.

Justin - You are talking about an analog circuit when you refer to detectors loosing up to 4 inches of depth. As explained in my question that happens when you are using a circuit to do the discrimination but the T2 does its discrimination by running a program on the processor and I don't think it does loose any depth using that method instead of a circuit to do the discrimination.

Now I done a little more research and although I have found out a little more it still doesn't really the question I an trying to get at. The piece of info I did find is that the All Metal mode uses a single filter method and the discrimination mode (at least in when using the DE process) used a double filter method. Now I don't know the exact effect this has on detecting in one mode or the other but this is what the manual says:

"The All Metal mode is more sensitive and offers better feel than the Discrimination mode, and
is used to find all metal objects present in the ground. The searchcoil must be in motion for
objects to be detected. This is a single filter search mode similar to the

 

[Elton]

The discrimination circuitry of a motion detector uses two channels of filtered signals to come up with a composite discrimination signal. The normal Discriminate signal is filtered to get rid of ground effects, but the resultant signal has lost its DC reference, and is now an AC signal. The only way to tell if it is good or bad is to filter the All Metal signal in identical circuits and compare its result to that of the discriminator. If both signals are the same, the discriminate target must be good, because an All Metal signal is "inherently good." If they are different, the Discriminate signal must be "bad.

 

[Frank in NH]

n/t

 

[Elton]

A metal detector uses a transmitting search coil inductively coupled to a receiving coil for detecting the presence of metal objects near the surface of the ground within the field of the coils. An oscillator generates a signal transmitted by the transmit coil, and the signals detected by the receive coil are coupled to the signal inputs of two synchronous demodulators. The output of the oscillator also is applied at different phases to the reference signal inputs of the two synchronous demodulators. The outputs of these demodulators then are passed through low pass and bandpass filters having a low cutoff frequency which is higher than the highest frequency components generated in the synchronous demodulators due to ground effects. The signals passing through the bandpass filters then are applied respectively to the signal input and reference signal input of a third synchronous demodulator, the output signal of which has an amplitude representative of the presence of metal objects and the polarity of which is an indication of the type of metal being detected. Undesired signals produced by ground effects are reduced by a considerable amount, and the output of the third synchronous demodulator is applied to a suitable indicator circuit

 

[SteveP(NH)]

The heart of my question is that all of the stuff in your previous 2 answers seems to describe different circuits that are used to process an analog signal. However isn't this stuff all done in software on a fully digital machine? In other words the signal coming in from the coil goes first through an analog to digital converter and is digitized into a stream of 1's and 0's and so things like its highest frequency is no longer an actual physical signal anylonger but is now just a different pattern of bits. The software that the processor is running does analysis on the bit stream that is the equivalent of the operations you describe. For example the output signal no longer has an amplitutude that directly represents the metal target any longer because that information has been digitized. The amplitude of the signal is either zero for off or 1 volt for on (i.e. 0 or 1) and so it can't be mixed with the original signal and fed into a 3rd demodulator.

Now a process equivalent to that can happen in the software program doing the signal analysis on the data stream but there isn't going to be any signal loss inherent in analog circuits.

This probably doesn't make any sense from an engineering perspective because I don't know enough about electrical engineering and digital signal analysis to frame my question in the correct technical terms. But what I am trying to get at is that as detectorists isn't a lot of our thinking about how a detector works based on what we learned about how analog designs work and this is changing now due to the emergence of all digital machines. I mean besides a machine like the T2 look at the Xterra 705. It has its analogue to digital converter in the coil so the signal is digitized before it even gets to the control box. What I am trying to understand is what effect does this new way of digitizing the signal before it is processed have on our understanding of the trade offs involved in using features such as AM versus discrim. Maybe the answer is none but I doubt it. For instance I don't think that running in discrim mode looses much if any depth on a digital machine.

And when I say digital machine, I don't just mean that it displays a target ID as a number instead of moving a needle on a meter but rather a machine that does all of its signal analysis in software and not by using demodulators and filters made of resistors and capicators and whatever else goes into one.

 

[SteveP(NH)]

n/t

 

[Elton]

A simple discrimination circuit measures the amount of distortion or shift and beeps or doesn't beep based on the settings of the machine. During the design phase of any metered-style machine, the engineer measures the amount of shift that the most common targets cause and programs a microprocessor to respond with a meter reading for those types of shifts.

The testing can include simple air tests, field tests in a controlled environment, such as a test garden, or even complex reports from several different field testers. But at some point, someone decides that a type of target shift represents a specific meter reading. While this information can give a detectorist a basis to dig or reject a target, it is in no way perfect.

 

[SteveP(NH)]

MY chain of thinking and questions came from a comment that Dave J made a while ago when responding to another post. I forget what the subject of the discussion was about but Dave made a comment something to the effect of now that discrimination is done in software it (the signal loss from discrimination circuits) doesn't really matter that much, but that is a whole different discussion. Then he went on to finish his answer about the original topic.

So all that stuff about how a discrimination circuit works doesn't really apply any longer since a computer program is now doing the same work that used to be done with electronic parts in a circuit. Or does it? That is what I am asking I guess. Or perhaps more to the point what do we as detectorists who started off using analog machines need to change in our thinking about when to use discrimination and when not to.

Don't feel dumb over this - I think this is new turf for us and I am not even sure I know enough about it myself to be able to frame my questions in an intelligent manner. I guess all this stuff is proprietary info and so and engineer or software engineer can't really jump in and straighten this out for us.

 

[Elton]

But in short..we can rely more on a Micro processors to determine dig or not dig than electronic circuits of the older machines......In My Opinion ..

Granted it will still allow targets of the same electronic footprint to give a good indication..... but I think it will reduce a lot of targets from being lumped in like on the older circuits... More defined areas of acceptance, or rejection..Can be determined by the Engineer who configures the micro chip set up. ?????????????

 

[SteveP(NH)]

I agree complete as I agree that target ID has gotten better over the last decade as more powerful processors, running more sophisticated signal analysis software does do a better job at target ID than designs from the late 90s did.

My questions I guess are more focused around what does this change from analog circuits to software mean for us when we are in the field and making choices to set up our machines. Here is a good for instance - cranking up the sensitivity on an older design usually meant loosing target ID accuracy but on a T2/F75 you get better target IDs (at the cost of more chatter). Are there other things like that which we haven't really incorporated into our set up and search routines?

 

[Monte]

SteveP, the detector design engineers have the task of making new detectors to meet the current demands and/or to serve the wishes of the owners and/or for the 'fun' of being creative and/or (in some cases) maybe just to keep their job. As we can see in reviewing all of the major detector makers just here in the USA (or those who purport to make all of their detectors here in the USA), we can see a broad range of detector offerings from mainly an 'analog' type design to perhaps the most technically 'digital' offering being produced today.

As you noted later in this thread, the reference to 'digital' is not confined to ONLY models that use an LCD type display rather than a needle meter type display. Heavens, not all needle designs worked alike, either! One of the things that amazes me is the alleged 'pressure' that some government regulations might be making on the detector design industry to eliminate lead or just the change in modern electronic equipment design that has gravitated toward a 'digital' influence.

But as we look back at the detector engineers we can see that there have been some pronounced changes. Some detector makers (engineers) might run a company but their personal skill set isn't up to par with the better design engineers. Other detector makers have made impressive steps to improve their competitive nature by bringing in new guys (and gals) to help design these 'modern' 'digital' designs, and that has left a lot of people (maybe including you), wondering if the new stuff is really better than the older detector offerings.

For me, I just rely on the often stated opinion (mine) that: "There is no such thing as a 'perfect' detector." There are quite a few that are very versatile, based upon myu methods of determining what performance abilities a detector provides, for the avid 'detectorist,' and many others that are also quite good for most 'average' metal detecting 'hobbyists.' Some of both offerings are in my personal detector arsenal for regular use.

Besides the obvious difference of AM accepting all targets and discrim not responding to targets which are being discriminated out what is the technical difference between the two modes.

Let's remember, too, that many detector models on the market have a control setting that is labeled *All Metal*, but it actually in NOT a true, Threshold-based or conventional All Metal mode. Worse yet, many of those models that are so labeled or described for search in the motion-based Discriminate moder do not actually accept ALL metal targets. This is easily seen in three of the four field-tests scenarios that I use to evaluate metal detectors.

To truly accept ALL metal targets, ferrous and non-ferrous, when in an "all metal accept" Discriminate mode in challenging ground they have to adjust low enough to accept the full range of target conductivities (often referred to as a true 180

 

[SteveP(NH)]

Actually I enjoyed reading your post quite a bit as it, for the most part, echos what I was trying to discuss when I first posted. Thanks for taking the time to post an answer. Hopefully more knowledgable about how these changes in the technology from analog to digital or perhaps it is better to say software based detectors change the way we should be thinking about detecting techniques (for a lack of a better way to put it).

You continue to enjoy the search too. I agree with you 100% on the use of multiple machines, I usually keep 3 different ones around myself.

 

[SteveP(NH)]

I was re-reading Dave J's whitepapers and came across this in one of them....

"There are many types of discriminators, all of which have some effect on

 

[jmario]

Wow excellent post I read it like ten times and really makes since .We used to call the unknown electronic part a Disgruntifacater weather it was in a radio,TV, or automobile