DEEPER PI DETECTION DEPTH

[Tinkerer]

Below is the TEM current wave form.

What is different?

About the same coil current and Flyback voltage.

A higher TX voltage and a much higher pulse repetition rate.

How does the target response compare? We will see the pictures later.

How does the power consumption compare? Big surprise, The TEM method uses a fraction of the power, because it recycles the power instead of burning it off in the damping resistor.

Tinkerer

[Midas]

OK well I guess that's enough semantics. Allow me to point out a practical difference between the TEM and true PI. Its recycling a great deal of the power which is great and allows you to use a reasonably low duty cycle of input power just like PI and get superior current flows. But the one component that doesn't get a rest that normally would, is the coil. At some point the amount of heat generated in the coil by the constantly flowing current will exceed what the coil can dissipate. If the coil is potted then I think that limit may come up sooner than you think.

If we instead truly pulse the coil at a low duty cycle then we should be able to generate far higher instantaneous current flows, and greater depths for the same average coil current.

I think what we want for the ultimate deep seeking metal detector is the recycling power of the TEM combined with the advantages of intermittent current flow that comes with PI.

[Tinkerer]

OK well I guess that's enough semantics. Allow me to point out a practical difference between the TEM and true PI. Its recycling a great deal of the power which is great and allows you to use a reasonably low duty cycle of input power just like PI and get superior current flows. But the one component that doesn't get a rest that normally would, is the coil. At some point the amount of heat generated in the coil by the constantly flowing current will exceed what the coil can dissipate. If the coil is potted then I think that limit may come up sooner than you think.

If we instead truly pulse the coil at a low duty cycle then we should be able to generate far higher instantaneous current flows, and greater depths for the same average coil current.

I think what we want for the ultimate deep seeking metal detector is the recycling power of the TEM combined with the advantages of intermittent current flow that comes with PI.

Hi Midas,

the simulations are not about semantics. I am trying to explain the different results by using one or the other method.
I posted the LTSpice file above, so that anyone can try all kinds of variations, but for the ones who do not have the time to face the learning curve, I show a few results.

I would also appreciate if some knowledgeable person would help improving the simulation.

You point out an important feature. If we adjust the traditional PI cycles to consume equal power as the TEM, what is the difference in the result?

I will show this with the simulations to come, but for now I can say that with the same power, the TEM generates a target response amplitude of about one magnitude higher.

Another very significant difference is the Signal to Noise ratio S/N. With the same power, we get also a ten times higher PPS, or Pulse Repetition Rate. This allows for integrating or averaging, 10 times more samples and that vastly improves the S/N.

Tinkerer

[Midas]

Hi Midas,

the simulations are not about semantics. I am trying to explain the different results by using one or the other method.
I posted the LTSpice file above, so that anyone can try all kinds of variations, but for the ones who do not have the time to face the learning curve, I show a few results.

I would also appreciate if some knowledgeable person would help improving the simulation.

You point out an important feature. If we adjust the traditional PI cycles to consume equal power as the TEM, what is the difference in the result?

I will show this with the simulations to come, but for now I can say that with the same power, the TEM generates a target response amplitude of about one magnitude higher.

Another very significant difference is the Signal to Noise ratio S/N. With the same power, we get also a ten times higher PPS, or Pulse Repetition Rate. This allows for integrating or averaging, 10 times more samples and that vastly improves the S/N.

Tinkerer

No, no it was me that was arguing the semantics not you. But its OK I'm done now, there's no point in discussing "PI or not PI' any further unless someone else wants to offer their opinion.

I do appreciate you posting the Sim results, its interesting stuff. It does seem like a technique with a lot of merit. I'm also wondering how the much the effciency suffers when you've got the coil over some highly mineralization ground. I'm guessing that's hard to simulate though.

[Tinkerer]

No, no it was me that was arguing the semantics not you. But its OK I'm done now, there's no point in discussing "PI or not PI' any further unless someone else wants to offer their opinion.

I do appreciate you posting the Sim results, its interesting stuff. It does seem like a technique with a lot of merit. I'm also wondering how the much the effciency suffers when you've got the coil over some highly mineralization ground. I'm guessing that's hard to simulate though.

Simulate highly mineralized ground, hmm, I wish I knew how to do that.
We know that it needs high power pulses to penetrate deep into mineralized ground, but the whole Ground Balance thing is much more complex as just a TX pulse.

Much to do with the RX. But, let's make a little test.

Set your coil on the ground, mild or hot, it does not matter.

Then take a good size junk of iron, (I use a 6 pound cannon ball) and set it in the center of the coil. (static hot soil) Without ground balancing or re-tuning, sweep a one gram gold target over the coil, at a distance. Can you detect the gold?

My detector has no problem with that. It still detects the gold at the same distance.

Tinkerer

[Midas]

Simulate highly mineralized ground, hmm, I wish I knew how to do that.
We know that it needs high power pulses to penetrate deep into mineralized ground, but the whole Ground Balance thing is much more complex as just a TX pulse.

Much to do with the RX. But, let's make a little test.

Set your coil on the ground, mild or hot, it does not matter.

Then take a good size junk of iron, (I use a 6 pound cannon ball) and set it in the center of the coil. (static hot soil) Without ground balancing or re-tuning, sweep a one gram gold target over the coil, at a distance. Can you detect the gold?

My detector has no problem with that. It still detects the gold at the same distance.

Tinkerer

Yep I get the same detecting depth on a 1g nugget regardless of any ground effect. I can't detect a 1g nugget at all. My current detector is a poor specimen. A new one based on Moodz's and some of your own work is in progress... slow progress.

[Midas]

Hi Tinkerer,

Sorry that was a little flippant. I am impressed that your able to do that. You must have achieved a wide dynamic range. Is the design very different to your published work? I assume its still IB and still uses differential amplifier.

Midas

[Tinkerer]

Hi Tinkerer,

Sorry that was a little flippant. I am impressed that your able to do that. You must have achieved a wide dynamic range. Is the design very different to your published work? I assume its still IB and still uses differential amplifier.

Midas

Hi Midas,

the design is different in many aspects, but it could be called an evolution from my earlier published work.
I still use a differential approach and an IB coil.

I wish you luck with your work in progress.

Tinkerer

[Tinkerer]

Below is the Moodz variation.
The coil current is the yellow trace. Take note of the peak amperes.
Not much difference in the circuit, but different output. The pulses are positive only.
The power consumption is only very slightly more.
We get more PPS, that is good.
I added the target response of the 5 targets of different size. Take note of the signal amplitude.
The peak target amplitude is 472mA

Tinkerer

[Midas]

You've got me confused now Tinkerer. How are you proposing to integrated the TEM method with Moodz's coil which obviously isn't IB ?

[Tinkerer]

You've got me confused now Tinkerer. How are you proposing to integrated the TEM method with Moodz's coil which obviously isn't IB ?

Don't ask me, I only made and show the simulation of Moodz's circuit inpost #12, above.

Tomorrow I will post the signal response of the TEM circuit, for comparison.

Tinkerer

[Midas]

Don't ask me, I only made and show the simulation of Moodz's circuit inpost #12, above.

Tomorrow I will post the signal response of the TEM circuit, for comparison.

Tinkerer

Oh I See. My bad, always leaping to the wrong conclusion. So is the only different between this sim and post #31 the diode?

Midas

[Tinkerer]

Oh I See. My bad, always leaping to the wrong conclusion. So is the only different between this sim and post #31 the diode?

Midas

The diode makes the big difference.
The sim has been adjusted for the power consumption to be about equal.

The induction is different.

The PPS is different.

Actually I got the TEM target response sim ready, so I post it here.



Below is the TEM method again.
Same power consumption, take note of the peak amperes of the yellow trace.
The pulses are bi-polar, this makes a big difference.
We get less PPS only about 4300 PPS. Still enough to do some integration to get a good S/N and keep the response fast.
Take note of the target signal amplitude. Same targets as the Moodz variation. These are 5 resistive targets like gold nuggets of different size.
You can see the peak target amplitude is over 6A
Tinkerer

[Aziz]

Below is the Moodz variation.
The coil current is the yellow trace. Take note of the peak amperes.
Not much difference in the circuit, but different output. The pulses are positive only.
The power consumption is only very slightly more.
We get more PPS, that is good.
I added the target response of the 5 targets of different size. Take note of the signal amplitude.
The peak target amplitude is 472mA

Tinkerer

Tinkerer,

that isn't the moodz's version (variation ok, why not). Connect the tuning capacitor to the anode of the diode (not cathode). You should get a sine like wave form at the resonance frequency.

Aziz

[moodz]

Tinkerer,

that isn't the moodz's version (variation ok, why not). Connect the tuning capacitor to the anode of the diode (not cathode). You should get a sine like wave form at the resonance frequency.

Aziz

Aziz is right mine did have a diode and the wave form is sinewave ... it is part of my new project D-IGSL ( Digital IGSL ) it rocks