automatic Balance - Eric?

[Anonymous]

Just found a free spectrum analyser for a PC
at:
http://polly.phys.msu.su/~zeld/oscill.html

 

[Anonymous]

Malcom, what multifrequency machines do is transmit a continuous rectangular-voltage waveform, which may be a simple square wave (CZ's), or a more complex waveform (Minelab, DFX). There is no transmit-off time such as there is in PI. The signal from the IB loop is demodulated either by extracting frequency-phase components (CZ's and DFX), or by using time-domain demodulators (which is apparently what Minelab does if their product corresponds to their patents, which it may not).
In principle, in either a PI or an MF machine, you could monitor and demodulate transmit current, and this has been done in the past for certain specialized purposes. However, getting good sensitivity is a bit of a problem, when you're trying to measure nanoamperes of target response buried in an ampere of coil current, which itself varies with temperature and with random fluctuations in power supply voltage.
I like your idea of doing a Fourier analysis on the PI receive waveform. In the future, heavy-duty computational techniques will come into more widespread use, as A/D converters and microprocessors get better and beeper engineers become more familiar with what can be done with them.
I think that Garrett does a baseband Fourier analysis on their so-called "imaging" model.
--Dave J.

 

[Anonymous]

Thanks Dave
I did have a balanced coil system in mind so
it would be a voltage waveform in both TX-on
and off. Would the TX-on current noise be as
much a problem with such a system?
Cheers
Malcolm

 

[Anonymous]

Malcolm:
You'll have the problems of loop imbalance to deal with. In a PI, these problems can be pretty rough, because the imbalance energy is concentrated mostly in that flyback spike.
That having been said-- yes, it is possible to just take the entire received signal and do a Fourier (or other) transform on it to extract information about what's happening in the field. The important thing to remember is that you must subtract the loop imbalance signals from the entire signal to get the target, which is a remainder. The imbalance signal may be thousands of times as strong as the signal from the target to be detected, so the problem of eliminating the imbalance signal is not a trivial one.
--Dave J.

 

[Anonymous]

Whatever method minelab use on their sovereign is in a word GOOD. And why because it works.
I am still for find anything iron in the disc mode, even with the disc knob on zero. If these sorts of results could be incorporated into a PI I may be inticed to buy one for coin hunting purposes.

 

[Anonymous]

I agree 100% with you that the Minelab sovereign is unbeatable in it's ability to reject all and any iron. I have NEVER seen any other detector even come close to it in this regard.
I use one with a Coiltek WOT coil. This coil is a work of art and gives the detector a lot of extra depth. My wife and I went hunting at a Rally in england last summer. We both used Soveriegns. While everyone around us was digging iron, neither of us dug a single piece of the stuff.
If anyone out there knows which Minelab patent explains the method used in the Sovereign, please let me know. It may not be the deepest detector but it's discrimination is indeed unbeatable.

 

[Anonymous]

The technology in the Sov probably corresponds roughly to US 5,537,041. I don't recall whether the patent mentions iron ID or not.
I suppose that anyone who fools around with multifrequency machines discovers fairly soon how to ID iron in the frequency/time domain. There's no one signal that does the trick like the reactive signal in a single freq. machine-- it's a matter of analyzing several different signals simultaneously. White's and Minelab have patents which describe modern iron ID systems.
Sorry, this is kind of a touchy subject, about which I probably shouldn't say any more.
--Dave J.

 

[Anonymous]

Malcolm,
I came up with this idea about the same time you did, going to try it next week.
I've got access to