signals within signals

Early on in the release of the MXT and Coinstrike (they came out within a week or two of each other) I compared undug targets using the stock coils. Overall the C$ did a much better job of proper ID whereas the MXT would average down good targets near iron. Later on after putting more time on the Coinstrike I came to feel that the reason for it being able to pick out and fairly accurately ID co-located targets was that rather than simply reporting a composite signal on co-located targets it was able to see a "signal with a signal" and report it. Good analog to digital converion of the received signal along with taking many looks at the digital signal "train" during analysis seemed to be what it was doing. My gut feeling is that this is beyond what is referred to as fast recovery... again the "signal within a signal". I can't state this unequivocally sinse I'm not a tech it is just my intuitive impression.


Tom



Edited 1 time(s). Last edit at 05/05/2009 09:10PM by Jackpine.

It is not uncommon for (high-end modern detectors) to try to report the highest ID received... during a certain 'span' time-frame. This is to say; If a detected target ..... while the coil is being swept....... wants to ID in the 30's ...then in the 50's... then in the 40's,,,,,,,,,, in a matter of a split second. NOW...... the electronics may take a 'selected' snapshot..... maybe a 80uS "window"....... and....within this 80uS window........ the detector electronics chooses the HIGHEST ""conductive"" ID (maybe not necessarily the highest signal strength... as in most cases).... and ,, thusly,,,,, report a higher ID.

Food for thought.

Thanks Tom I like your way of explaining the process. that is exactly what I am referring to.

I have to admit that at times it can be frustrating to try and get a split second high audio and ID to repeat. It takes a certain feel to comprehend the difference between a false and something that deserves a more critical look to see if the signal can be made to repeat in the same location. The sweet spot can be so small and the sweep length and angle so critical that one tends to perhaps give up a bit to easily at times. If it sounded too good don't give up!

Tom

That's exactly correct Tom. Yes, it takes skill to discern between the minute' falses..... and the small/tiny burp of a non-ferrous in co-locate w/iron.

Quote: "the detector electronics chooses the HIGHEST ""conductive"" ID (MAYBE NOT NECCESSARILY THE HIGHEST SIGNAL STRENGTH... AS IN MOST CASES" [emphasis added]

certainly this method is not always right but when it is, it is an eye opener.

Tom

Yes............ and there still remains "up-averaging".

Tom

IMO quite a few targets naturally up average regardless of the detector used. Many older analog detectors give higher than normal readings in some situations. On edge coins near iron are one particular example that comes to mind.

Cons; I do not know if this applies to the current Fisher detectors but one very popular model I used a couple years ago at times was prone to what I consider excessive up averaging of targets in neutral to mild ground. Deep in the clear targets became very loud, the TID on tabs read in the coin range and copper pennies read at the highest positive TID number. So a little more thought by the engineers is needed at times. Up averaging is fine for "hot ground" that tends to lower TID but the software also needs to recognize the interplay between a high yet stable sensitivity setting and what the ground analysis is seeing to prevent excessively high readings. If the state of the art cannot do this then there may be a need to allow the user to turn this function off. People do tend to complain when they dig pull tab after pull tab expecting a silver dime to appear and, if copper pennies read the highest TID then what is happening to the dimes and quarters?? Hmmmm

I am a strong proponent of up averaging but as in the case above it should have been field tested extensively under a wide variety of conditions and coil types before given the green light.


Tom

I had severe problems with the C$ up-averaging 'fringe-depth' targets..... including chewing gum foil......up into the 'coin' conductivity zone. There were plenty of areas where a tiny piece of chewing gum foil at 1"..... sounded/responded somewhat like a 10" deep coin. This was a show-stopper. Yet, the way the C$ would handle a certain type of bad dirt..... was to be HEAVILY commended.

Truthfully, I would rather a detector NOT up-average.... NOR down-average; yet, (thus far) it's nearly inevitable (side-effect). I also try NOT to become TOO fixated with VDI/ID numbers.