Triangular Wave Technology

[Aziz]

Can someone name a single metal detector design that will meet all of these goals?

Very trivial answer:
The PUD (or the UD)!

Aziz

[Carl-NC]

I think the Ultimate Detector has already become the Defunct Eternally Anticipated Detector.

[mikebg]

In a conventional PI it is the flyback pulse that produces usable eddy currents in the target. The eddy currents produced by the driving pulse are counterproductive.
For those who would like to play with the simulations, here is a RAR archive:
http://www.tb-electronic.de/pi_tech/CC1.rar
Happy simulating!
Thomas

Thomas, no need of simulations in time domain!
All know that we can't avoid the driving pulse. Then let we use frequency domain to show how to exploit it. The RX coil can't see the driving TX current. It sees only its derivative as AIR signal. Every competent designer should use induction balance to suppress AIR signal.
When you think that "The eddy currents produced by the driving pulse are counterproductive at conventional PI", then why you are still designing conventional PI instruments and still are using monocoil and time domain for SPICE analysis?

Start to design "WIDEBAND METAL DETECTORS". The term is from Frequency domain, where metal detectors are two types only: narrow band and wideband (MINELAB uses term "BROAD BAND": BBS, FBS, FBS2).
It is time all professional designers of "time domain" metal detectors to use Frequency domain for AC analysis and design.
Before to start special thread for AC analysis of WIDE (or BROAD band) metal detectors, I will pont a link for Jawa applet.

FOURIER SERIES APPLET
When "a" changes from 0 to 0.25 the signal changes from a square to trapezium and triangle wave.
You can control the applet to see how frequency components change.
http://cnyack.homestead.com/files/afourse/fssqtotr.htm

[Qiaozhi]

MikeBG - Are you proposing the use of SPICE frequency analysis in place of SPICE transient analysis?
If so, then you must understand the limitations:

SPICE frequency analysis is a "small-signal ac analysis". The models are linearized by the simulator before the frequency analysis starts. Since a PI detector is basically "large signal" (and highly nonlinear), then frequency analysis will give the wrong results. You mentioned Fourier analysis, which requires a transient analysis prior to running the FFT. The only way to avoid running a transient analysis is to use an harmonic balance algorithm to calculate the steady-state response. using this method you can simulate the non-linear aspects of the design. Alternatively you could use Cadence's Virtuoso simulator, which has a time-domain shooting Newton algorithm, to generate the frequency domain results. However, this is basically a souped-up transient method. In any case, these techniques are mainly used to speed up the simulation of highly nonlinear circuits, so that you can actually get to see the results before you reach retirement.

For PI simulation, you just cannot avoid using transient analysis, unless (of course) you only want to see the steady-state response. In which case, an harmonic balance simulation will be ok ... if you can afford it.

[PiTec]

Mike, the reason for posting the simulations is explained in the introduction of my posting. I agree with you that IB setups have a lot of advantages and that the standard PI TX waveform is not very effective, but there are also applications where it is difficult or impossible to use IB coils.

Thomas

[Prospector_Al]

Mike, the reason for posting the simulations is explained in the introduction of my posting. I agree with you that IB setups have a lot of advantages and that the standard PI TX waveform is not very effective, but there are also applications where it is difficult or impossible to use IB coils.

Thomas

HI Thomas,

I'm having trouble imagining situations where it would be impossible to use IB coils. Could you give some examples?

Cheers,

Prospector_Al

[Prospector_Al]

Hi Aziz,

Le Roi est mort, vive Le Roi! But who is the new King? A state secret?

Prospector_Al

[PiTec]

Hi Allan,

IB coils can become critical whenever they can’t be made rigid enough, e.g. large search loops. However, some additional electronic balancing may help in this case

Thomas

[Aziz]

Hi Allan,

IB coils can become critical whenever they can’t be made rigid enough, e.g. large search loops. However, some additional electronic balancing may help in this case

Thomas

Sounds like, I have to reveal another great invention to overcome this issue.
*LOL*
Aziz

[PiTec]

Hey, please leave something for us to invent Curious to know what you have in mind, of course. Maybe similar to what I am thinking about …

[Prospector_Al]

Sounds like, I have to reveal another great invention to overcome this issue.
*LOL*
Aziz

Hi Aziz,

Wie ist das Wetter in Krautland? Isn't it true that the imbalance of the coil system generates a signal that is identical to the "X" component of the ground signal. If the coil system does not flex continually it can therefore be balanced by the GB system, if its dynamic range is adequate.

But if you have another great invention up your sleeve, reveal it! Der König ist tot, lang lebe der König!

Allan

(Admin.: This seems to be an International Forum, so I hope you don't mind my excursions into foreign tongues..)

[Aziz]

Hi Aziz,

Wie ist das Wetter in Krautland? Isn't it true that the imbalance of the coil system generates a signal that is identical to the "X" component of the ground signal. If the coil system does not flex continually it can therefore be balanced by the GB system, if its dynamic range is adequate.

But if you have another great invention up your sleeve, reveal it! Der König ist tot, lang lebe der König!

Allan

(Admin.: This seems to be an International Forum, so I hope you don't mind my excursions into foreign tongues..)

Hi Allan,

I'm impressed. Very impressed. You must have pattern recognizing "Adleraugen" too.
Same solution.
Aziz

PS: The weather for the week: snow, cold, snow, cold, snow, cold, snow..

[Qiaozhi]

(Admin.: This seems to be an International Forum, so I hope you don't mind my excursions into foreign tongues..)

Although this is an international forum, the rules state: "I also ask that posts be written in English to maximize participation. If English is not your native language, just do the best you can."
However, excursions into another language are acceptable, as long as you also provide an English translation.
你明白吗？(Do you understand?)

[Aziz]

..
Der König ist tot, lang lebe der König!
..

Reminds me to the story of the "Des Kaisers neue Kleider" ("The Emperor's New Clothes").
Well, I can see the naked Kaiser.


Yep, "Der König ist tot, lang lebe der König!" ("The King is dead, long live the King!")

Aziz

[crane]

Hi Aziz,

Isn't it true that the imbalance of the coil system generates a signal that is identical to the "X" component of the ground signal. If the coil system does not flex continually it can therefore be balanced by the GB system, if its dynamic range is adequate.
Allan

To start with, Minelab detectors and those from Whites and other reputable firms all have sufficient dynamic range to handle >95% of the ground here in Australia. The rumour stating otherwise was started by you-know-who after a Minelab engineer pointed out an extremely rare occurrence with a GPX4500.

The only time X can be cancelled with the GB is when the window is centered over zero crossing but this won't work here on Oz gold fields because of the large R component.

Most coil designs won't false when struck on garden rocks such as sandstone but the same coil might be unusable if it flexes over mineralised ground. Coils such as the top hat and other bulky designs won't work here and if they did happen to give a measurable advantage then we would have to wind the gain back. The only real advantage in this case would be slightly better SN.

Perhaps you guys should take Eric's advise and get out of the house away from simulators and try your ideas in the real world for a change. It sometimes tricks me and I live here! What chance do you have?