What's new

Minelab GPZ8000 Deep Dive

geowizard

Supporting Vendor
🛒 Supporting Vendor
Joined
Apr 9, 2007
Messages
532
Reaction score
464
Golden Thread
0
Primary Interest:
All Treasure Hunting
New designs incorporate new "features" that can frustrate new users! :hello:
The result is in many cases a less than flattering review. Lets do a deep dive into the GPZ 8000 and discuss the technology and where that technology is found in Minelab predecessors.

- Geowizard
 
Last edited:
Based on the patents used in a new detector and the "prior art", we can make a few assumptions of what a new design uses that come from the past. :)

Minelab patents leading to the GPZ 8000: (source: Minelab)

GPZ 7000
US Patent Number
17/888081
D686924
9547065
8106770


GPX 6000
US Patent Number
11454736
9829598
11899156
9547065
11953642
8106770
10969511

PRODUCTS ON THIS PAGE
US Patent Number
10078148
8614576
11067715
10838103
7791345
9250348
9348053
11658416
19/302215
19/302164
11513252
12025766

- Geowizard
 
The 8k is a CCPI design, just like the 7k. It was my guess that they incorporated multi-period pulses but the ML engineer I talked to at the Wedderburn Jamboree thought it was still single period pulse. He was ME, not EE, so wasn't 100% sure. So until I put one on the oscope, I'll assume it's single period CCPI. In that case, it's identical to the 7k with these changes:
  • Lighter 6k mechanicals
  • New coil design
  • Almost certainly new/different RX processing algorithms
It suggests that if they offered these coils for the 7k along with a firmware update, a 7k could be updated into an 8k. But there's little profit in that.
 
Yes, I agree. The challenge of constant current has been in soils that affect the inductance of the transmitter coil. Current in the coil decreases with increasing reactance and impedance due to the equivalent of an "iron core" being added in the form of iron in the soil! :)

Patent 20180106925 addresses this issue.

Thanks for sharing!

- Geowizard
 
With reference to the potential of missing sunbakers;

Patent 20180106925 addresses this issue in paragraph [0020] going from page 1 to page 2. It discusses the issue with short time targets that are missed because they happen immediately after termination of the pulse and during the back emf (counter electromotive force) time.

This patent describes the value in capturing the high voltage cemf transient and regulating that voltage for use as a (free) constant, alternating high voltage source for the transmitter.

- Geowizard
 
Transmitter power:

Transmitter power in metal detectors is expressed in "Newtons". The calculation is a simple multiplication problem. Multiply the current (I) in amps times the Area (A) in square meters of the coil.

The formula for "moment" (M) in Newtons is M = I *A.

- Geowizard
 
Last edited:
Zonge NanoTem

Principles

To obtain very shallow information, the transmitted signal must go to zero very rapidly without ringing (oscillations of either the electronics or the wire loops themselves). Depending on loop characteristics, Zonge NanoTEM® transmitters turn off in approximately 1.5 microseconds.

This rapid turn-off, and the high-speed, analog-to-digital conversion, allows data collection at depths less than 2 meters and in areas with electrical resistivities in excess of 20,000 ohm-meters. The receiver records the decay curve as 31 windows (timegates), from approximately 1.5 microseconds after the transmitter turn-off to about 3 milliseconds.

 
This patent describes the value in capturing the high voltage cemf transient and regulating that voltage for use as a (free) constant, alternating high voltage source for the transmitter.
Energy recycling can be done in traditional PI designs, see US9366778. The usual way is to dump the flyback energy onto a cap which is used in the next turn-on phase. Often called a kickstart circuit. In the case of Minelab's CCPI, the flyback is captured on the parasitic capacitance of the coil itself.

Transmitter power in metal detectors is expressed in "Newtons".
Not ever. TX "power" (a misnomer in the first place) is expressed in ampere-turns (N*I) which has a unit of amperes, not Newtons. When it induces an eddy current in a target, then there is an opposing force between the fields that could be measured in Newtons, but you would only care about this if you were building an aluminum separator for a recycling center.
 
Which of the waveforms in US20180106925 are CCPI?
Or, alternatively, which apply to the 7k and 8k? (Same answer)
 
Here's the answer: Fig 2 and Fig 4 are CCPI waveforms, exactly what the GPZ7000 (and presumably the 8k) transmit. Fig2a shows the voltage with the high voltage spikes that drive the current change and the low "holding" voltages that maintain a constant current. Fig2b shows the coil current, essentially a square wave. You would sample for target and ground info after each transition. Even though coil current continues to flow it is DC and will not generate target eddies (ε = -dΦ/dt = 0) so the result is a purely resistive PI response. Ultra-low capacitance makes the transitions fast, but you could add capacitance to slow them down and even potentially sample during the transitions (requires an IB coil) to get a reactive response.

Fig7 is a traditional monopolar PI and Fig9 is a traditional bipolar PI (as used in the AF108). Fig8 is a monopolar pulse combined with a stepped pulse ─ this is what the GPX 6000 uses. Figs 7-9 only show the voltage waveforms; here is what the current waveform looks like in the GPX6k:
1775325780959.webp

So US20180106925 isn't really about CCPI or the 7k/8k. Rather, it's about all the various PI methods maintaining a constant peak current on every cycle to minimize ground noise.

The problem with using AI to analyze patents is that a patent is often not 100% self-contained. That is, it builds on prior knowledge (including prior patents) that may be difficult for AI to incorporate. In US20180106925, you need a prior understanding of the various PI methods that are presented. A patent also usually contains drawings that may not be so easy for AI to interpret. I've used AI to analyze patents but only as an initial step before I personally read the whole patent and figure it out.
 
There are several ways to do CCPI. The 7k basically uses a current square wave like this:
1775583999257.webp

Unlike traditional PI, there is no zero-current region for "listening" to the RX. But you don't need a zero-current region, all you need is a zero-current-change region and that is achieved by the "constant current" peaks, hence the term CCPI. Constant current means dI/dt = 0 and therefore dB/dt = 0, and it is a changing magnetic field that creates eddy currents. So the slews create target eddies, and the peaks are where we listen.

A bonus with this kind of CCPI is that the current peaks can be held with a relatively low voltage, perhaps as low as 1V, and if energy recycling is incorporated the current slews burn very little power. Overall, this can be far more fuel-efficient than a traditional total-loss PI and can even rival VLF. Another bonus is that it is naturally bipolar so EFE is canceled. And it is capable of being run over a wide range of frequencies.

The waveform above implies the slews are exponential but in an energy-recycling design (as with the 7k) the coil current dumps onto a capacitor so the slews are actually half-cosine:
1775585264410.webp

The resulting coil voltage is therefore a series of half-sine pulses:
1775585317379.webp

Peak voltage can be quite high. For a given TX inductance, the less capacitance the faster the slew and the higher the voltage.

The nature of CCPI can be confusing. The current is essentially a square wave, but some VLF detectors also drive the coil with a square wave. However, with VLF detectors we are talking about a square-wave voltage drive and with CCPI it's a square-wave current drive. Likewise, there is another PI design known as "half-sine," but this refers to a half-sine current whereas CCPI has a half-sine voltage.

CCPI can also create multi-period waveforms:
1775585748496.webp

My initial assumption was that the 8k would employ this, perhaps in a "fine mode" for small gold, but so far I see no evidence this is the case. And the Minelab (mechanical) engineer thought that the 8k had the same TX as the 7k, which was a single pulse rate. However, the 7k used 2 different pulse rates depending on modes: 1225Hz and 3700Hz. I would not assume the 8k is the same.
 
Lets do a deep dive into the GPZ 8000 and discuss the technology and where that technology is found in Minelab predecessors.
Wiz, I'm mostly posting this stuff for you. Is it making any sense? Is it better info than you're getting from AI?

Here's an interesting question: What, exactly, does the term "ZVT" really mean?
 
Metal detectors have many dimensions;

Like most things, there are applications that require larger size to cover a larger application! The same concepts apply to metal detecting on a small scale as they do on a large scale.

Treasure is in the eyes of the beholder! :)

- Geowizard
 
In a place in a far away galaxy, lived a young boy. His father was a Gold Prospector! :hello:

Gold prospectors often have friends that are dowsers. Such it was that a dowser came into the realm of prospecting with special tools called dowsing rods. Not to be judgemental and not to be easily convinced, the young lad moved on with dreams of someday being able to prospect for Gold using tools born out of technology. Technology has proven worthy giving us tools that elevate precious metals prospecting to previously unimaginable levels! The rest is history! Is the GPZ 8000(tm) the end game in detectors? :dontknow:

Stick around! There's more!

- Geowizard
 
Last edited:
Along comes the GPZ 8000(tm);

The subject of the GPZ 8000 has been given treatment here on TreasreNet and elsewhere. What remains to be seen is public perception and how nebulous (undefined) operator controls like depth and mineralization modes affect success or failure in the product. I am in the market to buy several deep sensing detectors. I the GPZ 8000 the solution? My Whites TM 808 can detect a Volkswagen at six feet! :)

What size of GOLD in troy ounces will it take for a GPZ 8000 to detect at six feet? Can a GPZ 8000 detect a Volkswagen at six feet? :dontknow:

- Geowizard
 
Transmitter power:

Transmitter power in metal detectors is expressed in "Newtons". The calculation is a simple multiplication problem. Multiply the current (I) in amps times the Area (A) in square meters of the coil.

The formula for "moment" (M) in Newtons is M = I *A.

- Geowizard
Newtons;

Newtons are Systeme International (SI) units for magnetic moment.

Moment is also described as Number of turns (N) x Current in Amps (I) x Area in meters (A).

- Geowizard
 
In a place in a far away galaxy, lived a young boy. His father was a Gold Prospector! :hello:

Gold prospectors often have friends that are dowsers. Such it was that a dowser came into the realm of prospecting with special tools called dowsing rods. Not to be judgemental and not to be easily convinced, the young lad moved on with dreams of someday being able to prospect for Gold using tools born out of technology. Technology has proven worthy giving us tools that elevate precious metals prospecting to previously unimaginable levels! The rest is history! Is the GPZ 8000(tm) the end game in detectors? :dontknow:

Stick around! There's more!

- Geowizard Win casino onlayn
Nice way to frame it - kind of shows the shift from old-school methods to tech-driven detecting pretty well
 

Users who are viewing this thread

Back
Top Bottom