Find's Treasure Forums

Welcome to Find's Treasure Forums, Guests!

You are viewing this forums as a guest which limits you to read only status.

Only registered members may post stories, questions, classifieds, reply to other posts, contact other members using built in messaging and use many other features found on these forums.

Why not register and join us today? It's free! (We don't share your email addresses with anyone.) We keep email addresses of our users to protect them and others from bad people posting things they shouldn't.

Click here to register!



Need Support Help?

Cannot log in?, click here to have new password emailed to you

1,2,3rd derivative Disc?

A

Anonymous

Guest
Are these forms of discrimination similar to signal processing techniques that yield the slope(1st deriv) and max. min.(2nd deriv) of a signal. What is it that a third derivative form of discrimination is doing in relation to metal detecting.
Thanks in advance.
John
 
Hi John,
I am not sure of your question and what type of discrimination you are referring to. If you are refering to the terminology used by Fisher on their VLF's where they call their discrimination filters, double derivatives, they are merely referring to the number of high pass filters used. Some VLF manufacturers have used 3 high pass filters so I guess they could loosly be called a 3 derivative machine.
I haven't programmed any DSP so I cannot tell you if the principles are the same, but I would suspect they would be. In a VLF, the high pass filters are designed to minimize frequencies below 2 to 3 hz which is the range of signal frequencies generated by the ground mineralization when sweeping a search coil.
In reality, VLF's use bandpass filters normally peaking somewhere around 10 Hz, but again, refer only to the number of high pass filters when explaining the number of filters they have in their machines. Having upper frequency cutoff filters would really make this type of detector a 4 filter machine, 2 low pass and 2 high pass.
I have asked one of the more noted designers of VLF machines, George Payne, for an explanation of filters as they are used in a VLF. In the most simple terms he explained that a typical RC time constant of a high pass filter used in a motion VLF detector usually ranges from 10 ms to 22 ms, depending upon the machine and the engineer. George did go on and explained that one company used a 3 filter design were the 3rd high pass filter had a much longer time constant. This was to further reduce the ground signal without seriously affecting the signal from a buried object which normally was much faster.
I guess as a very loose answer to your question would be the software written would, most likely be written to simulate the filter technology as it would be done using analog components. Obvously, more can be done digitally to simulate a brick wall type filter.
Again, I am not sure if I have even come close to answering your question, but did give it a go.
Reg
 
John,
A "high pass filter" when operating in its passband has a response proportional to the rate of change (slope) of the incoming signal. In other words, it is a differentiator and "calculates" the derivative of the incoming signal. All practical high pass filters "roll off" at some upper frequency, above which they are not differentiators. In typical metal detector practice, the rolloff frequency of a high pass filter is usually in the range of 6 to 20 Hz.
Differentiators (call them high pass filters if you like) are used in metal detectors to boost the target signals (which are predominantly in the 4-12 Hz range) relative to the signals from ground minerals, which are predominantly in the 1-3 Hz range.
Differentiate the signal once, and you have "autotune" or "SAT". Most gold prospecting machines do it this way. The response you hear to a metal target is a "zip" sound.
Differentiate it again, and you have "second derivative" or so-called "two-filter" response. This arrangement happens to be particularly well suited to discriminators. It is economical to manufacture and has good response characteristics. The sound is typically "bip".
Differentiate it a third time to further improve separation of target signal from ground interference, and you get a signal that's too confusing to listen to. But the signal can be sampled and used to operate a visual or tone ID system or a discrimination blanking system. In most such cases the signal you actually hear will be a second derivative signal.
A high pass filter with a rolloff has infinite bandwidth, and therefore infinite noise unless another rolloff (i.e., a low pass filter) is provided. In a typical op amp circuit-based inverting differentiator, the first rolloff (of the high pass filter) is done with a resistor in series with the input, and the low pass filter is done with a capacitor in the feedback network. The low pass filters are usually designed to roll off in the range of 8 to 30 Hz. Although the combination of a high pass and low pass filter comprises a very low-Q (nonresonant) bandpass filter, the underlying thinking and analysis is based on the concept of differentiation.
The underlying concept of the so-called "four-filter" machines is narrowband high-Q (resonant)bandpass filtering, although they may include high pass and low pass filters as well. The designation "four-filter" started with a particular model back in the early 1980's, and since that time has been applied to all bandpass-type designs regardless of the number of filters. When they were first introduced in the mid-1970's, the center frequency was usually around 12 Hz and the bandwidth was usually a lot less than that. If swept at just the right speed, "4-filter" machines give separation of target and ground signals slightly better than that which can be achieved by differentiation. Because their narrow bandwidth is associated with a long response time, the "four-filter" machines usually have rather smeared response and poor target separation. Some of the more recent "four filter" designs have broadened the bandwidth and lowered the center frequency to get quicker response and less dependence on sweep speed. These newer designs are usually more sensitive as well, because they "see" the big chunk of target signal that lies outside the passband of the earlier narrowband systems.
Since they first introduced their "motion discriminators" in the early 1980's, most of the Fisher, Tesoro, and Garrett machines have been second derivative types. Most of White's "motion discriminators" have been of the bandpass type, although their "Classic" series is second derivative type. I don't know enough about Minelab or the modern Bounty Hunter to comment with any confidence on what kind of motion discrimination systems they use. C-Scope in England used to do second derivative, but I'm not familiar with what they're doing these days.
As far as I know, all "real" gold prospecting machines offer first derivative all-metals operation. The recent White's GMT provides a type of second derivative operation as well. I'm under the impression that the new Fisher Gold Strike also provides second derivative operation. Some pulse induction machines use first derivative operation, primarily to eliminate drift but also to reduce residual ground noise.
--Dave Johnson
 
Top