A MD question 4 the tech-oriented

[eddiecurrent]

I read that a caveat for a detector that has t.I.d. is that the recovery speed suffers due to a lag resulting from the time it takes the circuit to process and analyse the target in order to ID it.
Any thoughts on this?

 

Amazon Forum Fav 👍

How to Research for Treasure Hunting and Metal Detecting From Lead Generation to Vetting - Grab it through Amazon!

[Terry Soloman]

Yeah. I think I'll have another beer.

 

[eddiecurrent]

So you regard that claim as ridiculous? Why?

 

[Jason in Enid]

I know your question was "claimed" to be true when the first TIDs were introduced. But look at the DEUS. It's one of the fastest recovering detectors on the market today, and not only is it using a TID, it's also wireless.

IMO, the claim is pure bunk.

 

[eddiecurrent]

So this claim has been floating around for a while?
I'm not an accomplished engineer, just a backyard tech, the idea I guess is that it may take..what..microseconds..milliseconds...for the incoming signal to be run thru the chips algorithym, whatever it might be, fourier domain analysis, I dunno, and then spit out the result as a number on the lcd screen...and compared with the time it might take for a swinging coil to pass from a nail to a coin...miliseconds?...the circuit investment wouldn't impact the recovery speed?
I don't know, but I like to learn things. I'll place you guys in the "amused skeptic" camp.
Thanks for your thoughts.

 

[Terry Soloman]

Eddie, think less and detect more. "Milliseconds," really?

 

[eddiecurrent]

Well, yeah. I think the time it takes for a coil to travel 3" could be in the high milliseconds.
I like to think and detect. ;-)

 

[Jason in Enid]

Well, yeah. I think the time it takes for a coil to travel 3" could be in the high milliseconds.
I like to think and detect. ;-)

You're swinging WAAAAY too fast!

 

[eddiecurrent]

You're swinging WAAAAY too fast!

So you've measured the time it takes for a coil to go 3"? (Divided by 2)
I'd estimate I swing about 5ft a second.
Btw, the fact Le-Deus is wireless, when you think about it, has no bearing or relevance to the recovery speed. The process has already taken place and the screen already bears the data.
Anyone else like to chime in?

 

[Jason in Enid]

So you've measured the time it takes for a coil to go 3"? (Divided by 2)
I'd estimate I swing about 5ft a second.
Btw, the fact Le-Deus is wireless, when you think about it, has no bearing or relevance to the recovery speed. The process has already taken place and the screen already bears the data.
Anyone else like to chime in?

Yes, I know how fast I swing. 4 seconds each direction is the normal speed, some times much slower than that. 3" in "milliseconds"? You do realize what that word means right?

 

[eddiecurrent]

I know what my guess feels like, I've never experimented to see what milliseconds seem like. Nor have I done the math, I'm just doing casual conversation. 5 ft a second is what I consider a med- slow swing.
My guess as to how long it takes for a circuit to do the ID is a guess too. It could be much longer than milliseconds.
Sometimes it takes my computer chess game as long as 2 minutes to think of a move. I have to complain at it. " C'mon! Make up your mind."

 

[Jason in Enid]

Milliseconds isn't a feeling, it's a measurement. One millisecond = 1/1000 of 1 second. A 5 second swing over 5 feet means that it takes you 1000 milliseconds per foot. So your 3 inches of movement requires 250/1000 seconds.

 

[eddiecurrent]

Yes, I know. But it takes me one second to cover 5 ft.. Then you have to divide by 2 (detect-analyse-reset-detect-analyse)

 

[Carl-NC]

There are two "speed" parameters that are often confused: latency, and recovery. Latency is how long it takes from the instance the coil is top dead center over the target, to when you hear the beep in your ears. This is dependent on a lot of factors: TX frequency, analog filters, digital filters, processor clock rate, processor loop speed, audio code & circuitry, and possible wireless link. For a stand-alone target, it doesn't matter so much on what the target is, or how deep it is.

Eddie, you're asking about recovery speed. This means, how close can I place 2 targets and get 2 distinct responses? This is dependent on what the targets are, how deep they are, and highly dependent on sweep speed. In addition to some of the factors that also affect latency. A comprehensive test would be long and impractical, so I've adopted a quick-n-easy test. I place a quarter between two 16d nails and sweep the coil exactly 3" over them, at exactly 3 ft per second; detector in disc mode, disc set to absolute min, iron grunt turned on if it has one. I start out with the nails far enough from the quarter to get 3 distinct beeps, then move the nails closer until I can no longer distinguish all 3 beeps. That is my number for "target separation," and even though it's not comprehensive, it's an apple-to-apple way of comparing machines.

In terms of detector design, what affects recovery speed the most are the analog and digital filters. These are the things that allow us to separate targets from ground mineralization. There is usually a trade-off between speed & depth, and in the past everyone wanted more depth. But now folks want better recovery. In my test, an average machine has about 4" of separation. At 3fps, this translates to 111 milliseconds. If you want 1" of separation (doable), that's 28ms. So you weren't so far off as your critics imagined.

Keep in mind that raw latency has little to do with recovery speed. A detector with a fast zip-zip response can have lousy separation, and a detector with slow mushy audio can have quite good separation. I know, I tested a whole bunch of them. Sometimes with utter surprise.

 

[eddiecurrent]

Thanks Carl, for sharing what you've learned. Interesting that there's that trade off between recovery speed and depth. I guess a similar principle explains loss of depth while in the disc mode.
I've observed that certain brands perform differently, a Tesoro for example loses more depth in disc mode than a Fisher. The processor clock speed makes sense too.

In audio, everything you add to the circuit between the signal input and the speaker output changes the fidelity of the sound, from a purists pov that can be bad, from someone else's pov that likes a little extra bas or treble, or likes to tailor it to the room, it can be good.

Depth vs information. I have noticed that some $400 machines perform as well as ones costing 3 times as much, I have a 1236 that's a beep&dig and it rivals my CZ in recovery speed and depth. That's without using the 2nd derivitive filter, of course.

What really suprised me was how the orientation of a nail next to a coin has such a dramatic effect on recovery and masking. If the nail is pointing toward the coin it fools the detector and requires more separation to unmask the coin, if it is vertically oriented, the effect is much less.

 

[ivan salis]

all metal mode would be the fastest and deepest mode --since there is no "sorting" needed -- simply "yep its metal" ...once sorting (type of metal) , depth of find and other factors have to be sorted by the machine ..it by the nature of the process's being done will be a bit slower --simply logical -- think if you are working (focused) solely on one thing vs multi tasking ...does the multi tasking slow you down? of course it does...divided effort

 

[Carl-NC]

all metal mode would be the fastest and deepest mode --since there is no "sorting" needed -- simply "yep its metal" ...once sorting (type of metal) , depth of find and other factors have to be sorted by the machine ..it by the nature of the process's being done will be a bit slower --simply logical -- think if you are working (focused) solely on one thing vs multi tasking ...does the multi tasking slow you down? of course it does...divided effort

Not so much. All metal is faster because it bypasses the target filters, which results in a faster signal. Calculating depth and target ID in modern micros is usually a sub-millisecond task.

 

[eddiecurrent]

Seems to be another reason to hunt in all metal, then flip to disc when you hit a target. The CZ's are described as being widescan while in auto-tune (motion all-metal) mode, giving a slightly larger detection field and a little more depth.

 

[Jackalope]

It seems that if a detector is designed to allow more reactivity to weak amplitude signals, it could be done by reducing the analog and digital filtering. But while you get a beep on a deep coin you also get many accompanying false beeps on ground minerals that have not been adequately filtered. I suppose it is just pragmatic for the detector companies to ensure their detector runs mostly quiet, avoiding all the chirps and blips from the ground matrix, even if there is a loss of sensitivity to weak targets (usually deeply buried targets). Don't want frustrated buyers returning used detectors after a few days of confusing background noise. About the nail thing: The big thing today is unmasking ability, at least with more seasoned detectorists. The horizontal sweep distance between a nail and a coin provides one indication of target separation ability. Of course, in the ground nails are haphazardly arranged, but for testing purposes it probably shows one standardized aspect worth measuring. A deep detector that nulls in a fields of nails isn't much use in a variety of areas, and sales are not going to be brisk if seasoned detectorists perceive it cannot unmask well. So, I think the new paradigm has driven the market toward better target separation - and yet, increasing effective depth is still a goal too (with increasing depth and improving unmasking being the holy grail). I've tested nails aplenty next to coins. Since iron has a permeability greater than one, the nail is a magnetic field conduit, with field lines crowding into the iron nail. Crassly, the nail is borrowing the primary field lines of the coil and redirecting them, which is bad news if the nails are pointing toward the coin, worse if the coin is between two nail directly pointing its way. The magnetic field lines would be emanating over and around the coin, which for most detectors makes the coin invisible (at least the non-ferrous signal). I'm sure Carl can correct this, but it seems to me likely the field lines are also more in parallel with the coin, and far less eddies are being driven in that situation than with the field lines interacting perpendicular to the coin surface (which it would if the coil's field was not being diverted by the iron nails to a more horizontal plane), thus reducing the eddy creation further (which is principle in driving the non-ferrous coin signal). Just an unsubstantiated postulate that seems to explain the nail effect. John

 

[eddiecurrent]

Very interesting. If this is the case then I can't imagine a way to counteract the phenomena. Permeability & Lines of force are lines of force...period.
Now...how about a theory on why fine gold chains are invisible.

My CZ's have a strange quirk I've formed a theoretical answer for...they give a squawk after you click them off. I think it's an inductive or transient spike formed in the coil (inductor) when the EM field collapses.