Concept Design

[ivconic]

If we focus solely on the requirements and criteria that Tony gave at the beginning; it is clear that the aim is to detect native gold and gold nuggets.
And here ML is far ahead of everyone else.
Otherwise, in other cases, with other targets; one can notice a lot of flaws or at least average behavior.
And I would have plenty of examples to cite where GPX didn't blow me away. But that's not the point.
When it comes to gold nuggets, it is definitely ahead of the rest. Even when the results are not breathtaking.
How did ML do it? The holy grail of all questions. But a lot can be guessed by reading their "white papers".
And there are several enthusiasts with very good measuring equipment, who have tried to analyze what is happening in ML coils.
I think that the main quality of ML detectors lies not in "how they did it" but "with what approach and what kind of material they did it".
First of all, ML products are made of super quality material. And a lot of attention was paid to improving the S/N ratio.
The company is rich, so it can afford the most advanced and latest technology. Exactly the opposite of what an average enthusiast and hobbyist can afford.
Trying to fully unravel "how they did it" is a complete waste of time. It's unnecessary.
And in the end, they already did it. Any further copying is just copying.
We should focus here on the basic guidelines that Tony has set. And to try to do the maximum with the material available to us.

[Tinkerer]

If we focus solely on the requirements and criteria that Tony gave at the beginning; it is clear that the aim is to detect native gold and gold nuggets.
And here ML is far ahead of everyone else.
Otherwise, in other cases, with other targets; one can notice a lot of flaws or at least average behavior.
And I would have plenty of examples to cite where GPX didn't blow me away. But that's not the point.
When it comes to gold nuggets, it is definitely ahead of the rest. Even when the results are not breathtaking.
How did ML do it? The holy grail of all questions. But a lot can be guessed by reading their "white papers".
And there are several enthusiasts with very good measuring equipment, who have tried to analyze what is happening in ML coils.
I think that the main quality of ML detectors lies not in "how they did it" but "with what approach and what kind of material they did it".
First of all, ML products are made of super quality material. And a lot of attention was paid to improving the S/N ratio.
The company is rich, so it can afford the most advanced and latest technology. Exactly the opposite of what an average enthusiast and hobbyist can afford.
Trying to fully unravel "how they did it" is a complete waste of time. It's unnecessary.
And in the end, they already did it. Any further copying is just copying.
We should focus here on the basic guidelines that Tony has set. And to try to do the maximum with the material available to us.

We should focus here on the basic guidelines that Tony has set. And to try to do the maximum with the material available to us.[/FONT][/QUOTE]

I agree that we have to do the best with the material available to us.
Our big advantage is that we can be creative. We can think of innovations and inventions. Among the forum members there are some very clever people who are capable of thinking outside the box.
The box is the existing limitations. The brick walls.
I strongly believe that we are capable of breaking through the brick walls.

[Willy Bayot]

Before going into physical implementations of electronic circuits or processing platforms, I would suggest we keep our discussion at the 'functional specs' level for a moment.

My question was :

What functions are still subject to improvements in functional operations of existing widely published PI systems?

For instance, about those features that we know very well are some weak points in PI technologies compared to VLF systems.
e.g discrimination, battery energy consumption, GB, ...

Is there still any potential progresses to be expected from a research project on these subjects?

[Orbit]

Before going into physical implementations of electronic circuits or processing platforms, I would suggest we keep our discussion at the 'functional specs' level for a moment.

My question was :

What functions are still subject to improvements in functional operations of existing widely published PI systems?

For instance, about those features that we know very well are some weak points in PI technologies compared to VLF systems.
e.g discrimination, battery energy consumption, GB, ...

Is there still any potential progresses to be expected from a research project on these subjects?

If we do discrimination like minelab we have weaknesses like those of vlf if we use discrimination like whites again we have some advantages but also disadvantages I tested all types of discrimination it's great for one thing, it's not good for another and vice versa..I'm not an expert here, I'm sure there are many better ones, I'm just persistent I work with a soldering iron I rarely use the simulator . And I can't express myself well because I don't know English ! I expect a big contribution from those who worked mainly Delta Pulse and Surf pi.

[ivconic]

Tony pardon my French; "CHIRP" stands for "SWEEP" as I understand.
In what manner would that sweep take place; linear or logarithmic.
Linear doesn't make much sense or need to me.
But a sweep between several different PPS makes sense to me.

[Tinkerer]

Tony pardon my French; "CHIRP" stands for "SWEEP" as I understand.
In what manner would that sweep take place; linear or logarithmic.
Linear doesn't make much sense or need to me.
But a sweep between several different PPS makes sense to me.

We want to have the highest PPS possible so we can integrate the maximum number of pulses. Higher integration = better S/N and doubling the number of samples integrated = double the signal amplitude.
Like 25K PPS is excellent for nuggets with a TC of 1us.
With a decay time of 40us, the eddy currents of large nuggets, let's say 200us TC will decay only a little and the rest of the eddy currents subtracts from the eddy currents of the next pulse.
With a cycle time of 200us, the eddy currents of such large targets decay by about 64% or 100usTC by about 85%.
So basically we want a mixture of short and long cycles maybe 30us, 40us, 50us, 100us, 200us and repeat. Or 30us 50us, 100us, 200us, 500us if we were searching for very large targets.
For very large targets we would use a large diameter coil, slower coil sweep speed.
For small targets we would use a small coil, like 200mm diameter with a coil sweep speed of 1m/s.

All this of course is impossible with a traditional PI, but perfectly doable with our Square Current Wave TX.

Don't confuse now the coil sweep speed with the CHIRP, (PPS sweep)

[Dean Sarelius]

Hi Gents - Just want to stimulate some thoughts on system design fundamentals.
Given the clock transition times of modern components and in order to reduce noise there are two design practices that should be considered.
Here are the two main ones that I follow below and please feel free to add to this list.

Number 1 - 4 Layer SMD PCB is mandatory, why because ground planes can be at much closer proximity to high switching currents providing a ground return for these currents isolating them from coupling into nearby circuits.
This will require the use of SMD components in order to maintain a close coupling to the ground plane.
If you wish to learn more on this subjest suggest to review Rick Hartley youtube live training link here https://www.youtube.com/watch?v=ySuUZEjARPY&t=2975s

Number 2 - Common Mode rejection on power rails. Given the large current pulses to the TX coil all efforts shoud be made to ensure that wherever possible common mode rejection can be maintained across power rails.
For example using negative regulators from the main supply rail as per Carl Morelands techniques instead of positive voltage regulators will ensure common mode rejection is maintained between the main supply rail and the lower supply rail.

[Tinkerer]

Hi Gents - Just want to stimulate some thoughts on system design fundamentals.
Given the clock transition times of modern components and in order to reduce noise there are two design practices that should be considered.
Here are the two main ones that I follow below and please feel free to add to this list.

Number 1 - 4 Layer SMD PCB is mandatory, why because ground planes can be at much closer proximity to high switching currents providing a ground return for these currents isolating them from coupling into nearby circuits.
This will require the use of SMD components in order to maintain a close coupling to the ground plane.
If you wish to learn more on this subjest suggest to review Rick Hartley youtube live training link here https://www.youtube.com/watch?v=ySuUZEjARPY&t=2975s

Number 2 - Common Mode rejection on power rails. Given the large current pulses to the TX coil all efforts shoud be made to ensure that wherever possible common mode rejection can be maintained across power rails.
For example using negative regulators from the main supply rail as per Carl Morelands techniques instead of positive voltage regulators will ensure common mode rejection is maintained between the main supply rail and the lower supply rail.

You bring up an important point I wanted to address:
The large current to the TX coil. Do we need that? NO. We can reduce the TX current by a factor of 2 or 3 or even more while still maintaining the same TX field strength.
Like everywhere else, compromises have to be made.
Weighing all the factors involved we can find the sweet spot.

[Willy Bayot]

You bring up an important point I wanted to address:
The large current to the TX coil. Do we need that? NO. We can reduce the TX current by a factor of 2 or 3 or even more while still maintaining the same TX field strength.
Like everywhere else, compromises have to be made.
Weighing all the factors involved we can find the sweet spot.

Indeed, the constant current XMIT system can easily accomodate with reduced currents (larger coil inductance) BUT beware about the HIGH SWITCHING FLYBACK voltages in the XMIT PCB (We speak about around1kV).
So, the layout and the components of the XMIT module should be VERY carefully defined.

[Carl-NC]

Continuing now with new material...

Based on everything brought up so far there is a clear inclination toward PI, and even CCPI. Further discussion should consider things like power consumption and discrimination, things that should have been explored in the Definition stage. Ferinstance, I know that iron is a problem in some areas, maybe not so much in other areas AFAIK. But a weakness with CCPI is that the reactive signal is incredibly short and not useful for iron ID. It's one area where the GPZ lags the GPX. And in any PI design you have to pay close attention to power consumption. That has barely been mentioned but is important. Finally, whatever we develop has to have excellent ground handling so think about how that would be done in any proposals.

So let's continue figuring out the high-level approach.

[KingJL]

... Based on everything brought up so far there is a clear inclination toward PI, and even CCPI...

I would recommend that we adopt the moodz alt2h-bridge as the TX. It can operate as bipolar CCPI, bipolar PI, bipolar SINE (including a hybrid sine/pusedo VLF mode, or VLF (in CC mode by adding a tuner cap to the RX coil to tune the RX to about 118% of TX CC frequency ). It is the most versatile TX that I have come across.

[Carl-NC]

That might be the most likely solution. For now we should be more high-level, and decide the best single approach to meeting the goals of the definition. I don't want to make this a do-everything-for-everybody design or it will never get done.

[Tinkerer]

Continuing now with new material...

Based on everything brought up so far there is a clear inclination toward PI, and even CCPI. Further discussion should consider things like power consumption and discrimination, things that should have been explored in the Definition stage. Ferinstance, I know that iron is a problem in some areas, maybe not so much in other areas AFAIK. But a weakness with CCPI is that the reactive signal is incredibly short and not useful for iron ID. It's one area where the GPZ lags the GPX. And in any PI design you have to pay close attention to power consumption. That has barely been mentioned but is important. Finally, whatever we develop has to have excellent ground handling so think about how that would be done in any proposals.

So let's continue figuring out the high-level approach.

But a weakness with CCPI is that the reactive signal is incredibly short and not useful for iron ID

I have been experimenting with demodulation of the reactive signal during the Flyback of the CCPI.
It is a matter now how to integrate this circuit into the RX.

Where should I post this schematic?

[Tinkerer]

I would recommend that we adopt the moodz alt2h-bridge as the TX. It can operate as bipolar CCPI, bipolar PI, bipolar SINE (including a hybrid sine/pusedo VLF mode, or VLF (in CC mode by adding a tuner cap to the RX coil to tune the RX to about 118% of TX CC frequency ). It is the most versatile TX that I have come across.

It is the most versatile TX that I have come across.
I agree.
Do you already have a PCB for that?

[Carl-NC]

Where should I post this schematic?

Probably a separate thread in this forum so it doesn't get lost.