Minelab GPZ 7000

[Davor]

Simulations also show global warming, yet there is no such thing going on for 18 years or so.

It is all in a patent. No need to dig any deeper.

[Monolith]

Some people still believe it is a PI

Is it a PI or not?

The simulation shows Flyback at about 270V

Target response:
1=Tau 500us
2= Tau 100us
3= Tau 10us
4=Tau 5us
5=Tau 1us

[Davor]

By stumbling upon a conversation on another forum I realised many people do believe GPZ is a PI, and seeing you froth about it I just had to throw a bone. No malice intended.

As you (almost) pointed before, GPZ is a step voltage device, not a PI. The obvious advantage of every step voltage IB device is that you may sample right away, and hence it is much better approach for detecting small targets, since PI is inherently deaf to those. There is also a disadvantage in sense that the X and R components bleed to each other due to the finite coil resistance, and the fact that there is a voltage change across a Tx coil during R sampling period. A certain manufacturer whose-name-must-not-be-said tried to fix this before using a negative resistance approach, and arrived - nowhere. Negative resistance is achieved by a device that utilises a noisy positive feedback loop, and instead of X bleeding in, they got noise instead.
GPZ fixes that by using a fixed low voltage source to compensate for resistive loss during R sampling period. To achieve momentary transition to steady current flow through a Tx coil during R sampling period, there must be a regulation loop that sets the high voltage properly. I have no idea if that's working any good in GPZ, but it is a clever way to go about it.
However, it simply is not a PI device. It is a step voltage IB, a VLF if you like, and a proof that there is a lot for VLF to catch up.

I also think most people throwing away their GPZ-s are simply those that are used to PI simplicity, cradled into the minds of OZ prospectors by the same manufacturer whose-name-must-not-be-said. It resembles a bit a lad that learns storks are not delivering babies - on his bachelors party.

I think knowing more is a good thing. Knowing X gives you discrimination, yet another thing OZ prospectors have no clue about. But moodz does. Guess giving the OZ prospectors a VDI would grant a manufacturer whose-name-must-not-be-said a permanent ban from Australia, so there you haven't got it.

I'm not sharing my method of killing X during R, not yet.

[crane]

Some people still believe it is a PI

Davor,
I was going to let this go but as you have decided to stir the pot.....

Anyone following this thread would know that Moodz said the ZED (not some unknown patent) has a fundamental problem in the feedback loop that prevents it being even better.

E.G., "( the GPZ is still a very good detector though ) ... it (obviously meaning the GPZ) has a fundamental problem in the feedback loop that prevents it being even better ..."
He now says he was referring to one of many patents and/or control loops???

Why claim to have a superior method when it seems the hypothesis hasn't even been tested yet???

As to "some people still believe it is a PI". A pi doesn't need to address the X component and yet it is clearly audible in the following link and Moodz seems to think this flaw is an achievement?? Is his method PI??

https://www.youtube.com/watch?v=UHBV2hZrivo

The pulse in conventional PI is the magnetic step that occurs when the low voltage applied to the coil is suddenly switched off, ie , at T=0.
The GPZ magnetic step also occurs at T=0.

The main difference is that conventional PI ignores the X component because it is only present during transmission, ie, during the low voltage period and the spike.

[Davor]

Some people still believe it is a PI

[moodz]

Not sure what I mean??? Very simple really because you say the ZED has a fundamental problem in the feedback loop and yet we know it does an exceptional job of nulling the X component and that there is nothing in the public domain to indicate precisely how this is being done, which makes me wonder why you say it has a problem??

1. Yes, the oOo configuration is unconventional but where have you, Deemon or anyone else here mentioned why this configuration is needed and how did you discover this problem and how did you conclude it could be solved with one of the configurations shown in the patent?? And at the end of the day you still need a method to null the very large VRM ground signal. The VRM problem isn't a new problem so what do you propose??

2. A control loop must pass at final circuit gains or it fails, surely this is obvious??

3. "If they don't use that circuit topology then ... the technology is not patented ??" You lost me and a thousand others with that one!!!

hmm ... you seem to be taking this very personally ...

You first implied the coil was conventional .. it is not ( however it is prior art topology ) ... and it would help with ground balance.

Control loops rarely pick off the last gain point except in very simple designs ..... nested or modular control loops are usually used. eg the feedback path on an opamp is a control loop but localised generally to the single opamp.

"You" said they dont use the circuits in the patent above .. I was questioning your incorrect statement ... please read some more patents ... patents dont usually have the actual circuit ... they have a functional description / diagram ( eg a switch connected to a coil etc etc ) which describes an embodiment of the invention that can be "translated" to an actual device by someone skilled in the art ... you need to skilled in the art though. :-)

The GPZ is covered by a number of patents ... the patents do describe some of the key functions of the GPZ.

Finally I never said the GPZ had a problem with X / VRM ... I said the control loop described in the patent has a problem ... who said there is only one control loop and which one anyway ?? nuff said.

[crane]

Not sure what you mean ....

1. Its not really a conventional coil ... oOo confguration
2. Control loops lose dynamic range and stability if the gain is too high... the ground and X must be removed prior to final gains.
3. If they dont use that circuit topolgy then ... the technology is not patented ??

Not sure what I mean??? Very simple really because you say the ZED has a fundamental problem in the feedback loop and yet we know it does an exceptional job of nulling the X component and that there is nothing in the public domain to indicate precisely how this is being done, which makes me wonder why you say it has a problem??

1. Yes, the oOo configuration is unconventional but where have you, Deemon or anyone else here mentioned why this configuration is needed and how did you discover this problem and how did you conclude it could be solved with one of the configurations shown in the patent?? And at the end of the day you still need a method to null the very large VRM ground signal. The VRM problem isn't a new problem so what do you propose??

2. A control loop must pass at final circuit gains or it fails, surely this is obvious??

3. "If they don't use that circuit topology then ... the technology is not patented ??" You lost me and a thousand others with that one!!!

[moodz]

It would be interesting to know how you came to this conclusion knowing that the ZED does an exceptional job at nulling the X component and that ML don't use the crude circuits shown in their patents??

As to someone here building something better, note that after figuring out how to obtain a perfect X null, and figuring out how to address the unusual problems with conventional coils, they still have to devise a method that will null the very large VRM ground signal and hopefully do it as efficiently as ML. This last bit alone would be a challenge for most!

And note that the VRM and X signals must be nulled at very high final circuit gains, not at the low gains seen at the preamp output.

Not sure what you mean ....

1. Its not really a conventional coil ... oOo confguration
2. Control loops lose dynamic range and stability if the gain is too high... the ground and X must be removed prior to final gains.
3. If they dont use that circuit topolgy then ... the technology is not patented ??

[crane]

... it has a fundamental problem in the feedback loop that prevents it being even better ... more info after I manufacture and test the first board.

moodz

It would be interesting to know how you came to this conclusion knowing that the ZED does an exceptional job at nulling the X component and that ML don't use the crude circuits shown in their patents??

As to someone here building something better, note that after figuring out how to obtain a perfect X null, and figuring out how to address the unusual problems with conventional coils, they still have to devise a method that will null the very large VRM ground signal and hopefully do it as efficiently as ML. This last bit alone would be a challenge for most!

And note that the VRM and X signals must be nulled at very high final circuit gains, not at the low gains seen at the preamp output.

[Ronk]

nice post thankyou Kev nice cro too

[Tinkerer]

http://www.detectorprospector.com/fo...elab-gpz-7000/

Steve Herschbach quotes "6000 feet" ... not meters .... seems reasonable.

You can see that the TX coil is indeed Litzwire.

[moodz]

So Kev, who told you that story ?

http://www.detectorprospector.com/forum/topic/670-gpz-14-coil-for-minelab-gpz-7000/

Steve Herschbach quotes "6000 feet" ... not meters .... seems reasonable.

[moodz]

... http://www.minelab.com/__files/f/266...Technology.pdf

this paper is reasonable and there is some important info here ... the amount of wire and battery life is secondary ..

Deemon is on the right track .. it is most likely probable that a much better detector than even the GPZ can be made ( the GPZ is still a very good detector though ) ... it has a fundamental problem in the feedback loop that prevents it being even better ... more info after I manufacture and test the first board.

moodz

[Tinkerer]

I would not dispute that as it could be the total length of the strands of wire in the Litz wire in either coil.

Regards, Ian.

That seems to be about right.
I saw some Utube of the exposed coil wires and wave forms. It must be out there to find for anyone interested.

[IBGold]

I would not dispute that as it could be the total length of the strands of wire in the Litz wire in either coil.

Regards, Ian.