Zero Voltage Transmission (ZVT)

[jasong]

I spent some time reading the US public patent document last night and I think I have a very basic idea and perhaps not fully correct of how that coil works now. They use one (or both, the patent isn't clear) of the D's to monitor changes in the TX coil due to varying ground permeability aka mineralization and then use a negative feedback amp to keep the current constant in the TX since the reactance of the coil changes as you move over different ground or targets. So that basically ground noise and effects are nulled out constantly at what the monitor sample rate is. I think that's also where they get the term Zero Voltage Transmission since V = 0 when absolute value of di/dt > 0, but I'm not sure about that and abs(di/dt) > 0 is only true when speaking in terms of averages which they do explain. I can't figure out if I am missing something on understanding the bigger ZVT picture though or if it's kinda a semantics thing? You can see in the provided sample waveforms there is always a non-zero current present which differs from traditional PI designs.

The other net effect is that the coil remains in a critically damped state more or less constantly rather than changing as you swing and thus interefering with the decay time and thus the RX sampling period as with a traditional PI - a simple way to think about that is if your coil is still ringing during RX sampling then you have a hard time listening for target responses. It seems like that is where they are able to sample and detect lower conductivity or quick decay rate gold targets maybe?

So it's kind of a mono coil but not, and kind of a DD coil but not. Looking at the sample waveforms and their description it looks like a PI in theory still, but kinda not too.

Also, there is a point made about programs and updates in that document. Which leads to some interesting speculations about future models...maybe just program updates? No idea at all, just guessing.

Am I understanding all this correctly or have I steered awry somewhere? Getting through the patent legalese speak is kinda tough in that document for me.

[Steve Herschbach]

I am the first to admit there are lots of people smarter than me when it comes to the technical end and especially trying to make sense of patents. They are often very hard to decipher on purpose even for people that know the stuff.

What I am saying jasong is I am guessing you have a better handle on it than I do. There is a point where my eyes glaze over and I just focus on what it actually does as opposed to what is under the hood.

I think the intricacies of the way detectors work must be every bit or even more fascinating than detecting itself for certain personality types. It seems like most of the really brilliant people on the tech side do little or no detecting. And many of the best detectorists could care less about how the detector performs the magic. It is a rare person that is both a great detectorist and also has a superb grasp on the technology behind it. I think Jonathan Porter is one of those people.

Me, I just love prospecting and detecting. Learning the tech and yakking about it is an extension of that for when I can't actually be in the field, and it does help to know a bit about why things work the way they do.

But anyone that can read those patents and fully understand them I am in awe of. My brain just does not work like that!

Analogies and simplifications work best for me. Simplest one I have heard is that ZVT is like the detecting equivalent of Dolby for audio. The noise is removed so the signal is clearer. I think half of Candy's genius is his melding of audio and detecting technology signal processing techniques as he is proficient in both.

 

 

EDIT ADDED 3/26/2015 Bruce Candy Technical Paper - GPZ 7000 Zero Voltage Transmission (ZVT) Explained

[Reno Chris]

monitor changes in the TX coil due to varying ground permeability aka mineralization and then use a negative feedback amp to keep the current constant in the TX since the reactance of the coil changes as you move over different ground or targets. So that basically ground noise and effects are nulled out constantly at what the monitor sample rate is. I think that's also where they get the term Zero Voltage Transmission since V = 0 when absolute value of di/dt > 0, but I'm not sure about that and abs(di/dt) > 0 is only true when speaking in terms of averages which they do explain. I can't figure out if I am missing something on understanding the bigger ZVT picture though or if it's kinda a semantics thing? You can see in the provided sample waveforms there is always a non-zero current present which differs from traditional PI designs.

OK, I understand the concept of change in soil reactantance as the coil moves across the surface and how they could use a feedback loop and keep the current (i) constant. That makes sense. However since I = V/R, if V = 0, then I must also equal zero by simple algebra. if V = 0, then I = 0 and the absolute value of di/dt equals zero. so I understand how they could call it a constant current detector, but still don't get the term zero voltage transmission, as the simple interpretation of "zero voltage transmission" means zero TX and no voltage or current goes into the ground. Perhaps because the polarity of the system is reversed back and forth with a fairly high frequency, and a high level of control, perhaps they're claiming a net zero voltage and current because of the frequent reversals? Like the sum of the all values plus and minus equals zero? I really have no idea where the zero voltage transmission term derives.

If you have not watched the introductory video, do so and pay extra close attention to the segment where Bruce Candy has an open coil and talks about its operation (screenshot above). The center coil of the DOD configuration is the transmit coil and the video shows it to be made of Litz wire and have a number of windings comparable to other coils that I have seen opened in the past. What the uH rating is for this coil I have no idea, but the inductance is going to be in the ballpark of similar coils used on past metal detectors. It is the to side receive coils that are of great interest to me.

You can see in the video that these coils are made of a very fine gauge copper wire wound into a huge number of windings in each coil. This is a huge amount of copper even with the fine gauge of wire used in the coil, this goes a long way to explain why the coil seems so heavy - that's a lot of copper. Because these coils have so many, many windings, they will have a large inductance and be very sensitive to detecting any changes in magnetic fields beneath them. I think this unusual design and the very high inductance in the receive coils is an important part of the new technology and the obvious explanation of why this design cannot be simply adapted to existing GPX detectors.

For years prospectors have speculated about more power being output to the transmit part of the coil and how that might increase sensitivity of deep targets. Of course the concept comes with its own set of problems. The other possibility for increased depth sensitivity is to greatly increase the sensitivity of the receive coil by adding a lot more windings to the coil. This also comes with its own set of problems. Apparently minelab has, in the GPZ 7000, worked out a way to resolve the problems associated with a much higher inductance receive coil or coils.

[klunker]

  Thanks guys . Even someone as thick headed as me can get bits of useful information from your discussions. I believe it's more important than many realize to have at least a crude understanding of what a detector is doing and why.

[jasong]

OK, I understand the concept of change in soil reactantance as the coil moves across the surface and how they could use a feedback loop and keep the current (i) constant. That makes sense. However since I = V/R, if V = 0, then I must also equal zero by simple algebra. if V = 0, then I = 0 and the absolute value of di/dt equals zero. 

 

If we're just dealing with DC then that would be true but even with a pulse you have reactant components to the coil since delta functions (perfect impulses) only exist on paper. The coil has both inductive and capacitive properties that are not constant and can change due to a lot of different factors. The mag field also rings to some extent which provides another AC component to consider, though a big part of their invention I believe is to reduce this effect so that the coil is always in a critically damped state, but even then we still have AC components since it the coil still has a finite decay time (it's physically impossible not to, I believe).

 

So even in the case of a pulse we still have to look at the entire system from an AC perspective, and voltage and current can be out of phase with each over an inductor in AC and voltage can be zero when the current is constant.

 

That's where you get V = L*(di/dt). di/dt is the derivative of the current with respect to time, in other words - what is the rate of change of current over time. I ignored L (inductance) for simplicity sake since it doesn't change the real important part of it all, which is that when current is not changing, ie has a rate of change equal to zero, then voltage must be zero.

 

So, that's why in the sample waveforms you see the current staying at some non-zero value at all points during the operation of the detector.

 

Another way to look at this, which to me was one of the most interesting things I ever learned in school, is from a Fourier perspective. Any pulse, and even more generally any waveform (even a square wave) can be broken down and described in terms of an infinite summation of sine wave of varying frequencies. I can almost guarantee part of the programming in the GPX and GPZ uses Fast Fourier Transforms to analyze the RX signal, a lot of MCU have the capability built right into the chips these days.

[jasong]

 

So, that's why in the sample waveforms you see the current staying at some non-zero value at all points during the operation of the detector.

 

 

What I also meant to explain there was that even though the current is always non-zero with the GPZ, the voltage itself can be zero and probably is only non-zero when the pulse itself has some sort of modulation (some series of pulses, these are kinda like the timings we know) rather than a just the constant current portion of the waveform. Which is probably also why we still have the same timings. And makes me wonder if we might see timing upgrades via program updates rather than new physical models, but that is 100% speculation on my part.

[wirelessguy]

Jasong, can you provide the patent number you are referring to please?  Any other GPZ 7000 relevant patents numbers appreciated as well.  thx

[eurodigger]

Thanks for the replys, little to technical for me to grasp fully, but its enough for me to work out this isnt a GPX5000 with a different coil and box.

 

Do you think the new coil design is the reason there is no discrimination option with the detector?

[Steve Herschbach]

Constant current metal detector with a driven transmit coil http://www.google.com/patents/US20100141247?cl=en

[wirelessguy]

Thanks for the link Steve.