PI signal snapshots

[Paul]

Carl,

I think your understanding of Electrodynamics is mixed up, lol. Sorry, I couldn't resist. So I lied ... just thought you'd you like see something. I'm sure Reg is your buddy, but I don't really appreciate taking sides in blind manner. Your admin., block my IP if you want. I stick up for what's right. If you want carry on your blind sightedness then take a look at post 27, this thread. The quote was exact. Can't imagine anyone wanting to be called mixed up.

Say what you want, have the last say. Facts will always remain facts.

Paul

[Carl-NC]

Paul, this thread was a wonderful, detailed discussion on PI, with good info posted by both you and Reg, but you seem to have taken his comment--I think you are getting magnetic properties mixed up with properties that affect the decay of the signal--as something personal, when it was not meant to be so. I have no desire or intention to ban anyone, I'm only trying to maintain civility.

Thanks...

[Paul]

Very well, not sure what to say on that except it was Reg who ended the conversation by the mere fact that I voiced my dislike for calling me "mixed up."

...

Carl, is it true that the transmitting coil is an issue. I see their issues and have the obvious solution, but was wondering if you also knew the solution.

Paul

[Carl-NC]

Carl, is it true that the transmitting coil is an issue. I see their issues and have the obvious solution, but was wondering if you also knew the solution.

Are you asking about PI coils in general? If so then, yes, there are techniques to produce a better TX coil. Most center around the desire to minimize parasitic capacitance, to enable faster responses and faster sampling. Maybe you could elaborate?

- Carl

[robby_h]

Paul,

Not true?
You can easily see the huge effect that a moderate sample of typical ironstone ground has in picture (3). If I was to instead present a very much larger sample of magnetite (inches thick and much larger than the dia of the coil) to the coil in air as shown in picture (2) then you would only get a response similar to (1) and this response would be due to the coil's increase in inductance, not magnetic lag. If the coil was fully damped (it isn't) and has no late time energy still being released into the system from some capacitance then you would only see a replica of the same curve to zero shifted a little bit further back in time without the small late time apparent lag seen early in (1). Pure magnetite is classed as having an extremely high susceptibility and extremely low remnance.
Your chunk, if pure magnetite, should not show any "true" magnetic lag in a pi but will only affect the coil's properties resulting in a curve very similar to that from a very short TC non-ferrous object such as a bit of thin section al-foil.

[robby_h]

Hi Reg,
The gain was reduced a lot, to around 20 if I remember correctly. The opamp used is the old 5534. The gains at 120 show very large shifts for metal at the preamp output but unfortunately the ground takes over. Good on beachsand though. I think there is some evidence emerging from these tests that shows Minelab's tx method is more likely to suffer from near magnetic saturation than the methods Eric favours. It is very difficult as we can't physically see these domains and have to rely on outcomes from tests that may in themselves be somewhat flawed.

[Paul]

Your chunk, if pure magnetite, should not show any "true" magnetic lag in a pi

Robby_h,
All magnetic materials have lag. To not have any lag would require perfect material with zero impurities with zero friction. Such imperfections cause resistance which is manifested as heat. Additionally it would require that light travel at infinite velocity. Given light travels at c this manifests as radiation resistance, which is wasted energy given off as far field Electromagnetic waves.

Furthermore, the lag is determined not only by material type, but size and shape. A large piece of magnetite will surely generate a lag that extends far beyond 1 usec. It's simple, the large the piece, the more lag. Take a rock size magnetite and you'll get over 10 usec lag.

Peace,
Paul

[Ferric Toes]

Magnetite.

Oh how I wish I still had all the magnetite that was in my possession till a few years ago. All I have left out of around 50lbs, is a mere 4oz lump. I dumped the rest when I moved from Oxford to Devon, along with many other things that I wish I still had. All this magnetite came from the Kiruna iron mine in Northern Sweden and is about 60% pure. What was I doing there? Why, installing a PI metal detector on one of their conveyor belts.

The main problem there was due to the high permiability of the magnetite, in that it distorted the field from the search coils in a variable way, depending on the depth of material on the conveyor. The steel structure of the conveyor, which is fixed in position relative to the coils and which can normally be balanced out by a fixed offset in the electronics, now gave changing signals. This was later cured by a different coil design, and was the reason I obtained a considerable quantity of this magnetite for experimental purposes.

At no time did I observe any magnetic lag, or viscosity effects, from this material, even with large rock size pieces. The piece I still have, I use for test purposes, and with a very small coil to give a concentrated field, no lag is observable. That is not to say that there isn't any; just that with the present equipment, it can't be observed or measured. This "equipment" includes a PI that can sample down to a microsecond after the TX off. With that device I do see a good signal which decays into the noise inside of 10uS, but that is due to its conductivity, not viscosity. A dc test meter with probes placed 1/2in apart on the magnetite reads about 100 ohms.

The ferrite rods that I use for probe coils behave the same. With the coil wound tightly on the rod, no viscous signal. Get the wrong grade of ferrite though, and you can clearly see the difference.

Oxidised magnetite, or maghemite, is quite a different story.

Ferric Toes

[Paul]

no lag is observable. That is not to say that there isn't any; just that with the present equipment, it can't be observed or measured. This "equipment" includes a PI that can sample down to a microsecond after the TX off.

I have a legitimate piece of magnetite that was authenticated and purchased from a geologist at a store. It is roughly 1.23" x 1.63 x 1.65. Less than a 1 Gauss field generating coil quickly released generates a lag extending beyond 1 usec.

Some possibilities could be that your PI isn't generating a strong enough field, or the collapsing current di/dt is not sufficient, or the coil and circuitry are not meant to collapse and measure that quickly or perhaps the material was hematite and similar. I am not aware of any PI's coming from large corporations that allow and advertise their units can accurately work down to 1 usec. There seems to be a ton of smaller companies and some may actually allow such adjustments, but they probably would not work under 10 usec.

I witnessed the lag on an oscilloscope over one month ago. Who knows, maybe I have super magnetite.

There could be a host of reasons why a certain PI unit doesn't report such lag from large magnetite. Other possibilities could be some GB cancellations or perhaps it's doing a good job at blocking magnetic materials.

Best method to really verify it is just get sufficient coil, generate sufficient field and quickly collapse it. Not difficult so far. Then view the output from a small coil on scope. View with and without the magnetite.

Also, make sure it's magnetite, not maghemite and not hematite. Big difference. Who knows, maybe I got ripped off and what I really have is maghemite, lol. Not sure except that I paid 30 bucks for it! Darn geologist! They're the greatest.

With that device I do see a good signal which decays into the noise inside of 10uS, but that is due to its conductivity, not viscosity.

Ahh, you're doing it wrong. You need to look at it exactly when the APPLIED magnetic field collapses. Two things occur when the current completely halts. 1. All remaining free magnetic fields collapse and convert into far fields, which collapses into the coil and continues to traverse through space. This takes not much longer than a few nanosec for say a three-foot coil. 2. The magnetic material lags caused by self-generating magnetic fields within the magnetic material itself. This simply depends on material type, size, and shape.

Get the wrong grade of ferrite though, and you can clearly see the difference.

:-) Now are you talking about natural magnetite or some commercial grade ferrite rod? Both contain lag, but commercial grades are usually light years better than natural stuff. Good point on your part anyways. Natural magnetite should vary. Just how much I'm not certain. Like I said, may be I have some rare super magnetite. Gosh, perhaps $30 was a bargain, lol.

Oxidised magnetite, or maghemite, is quite a different story.

Yes, but we're all interested in them since they're all natural and common. I know maghemite is common in good old California. Magnetite and Hematite should be common everywhere.

Paul

[Ferric Toes]

I have a legitimate piece of magnetite that was authenticated and purchased from a geologist at a store. It is roughly 1.23" x 1.63 x 1.65. Less than a 1 Gauss field generating coil quickly released generates a lag extending beyond 1 usec.
Paul

Much natural magnetite has maghemite associated with it. Much of the literature on the subject mentions the susceptibility/viscosity ratio of soils and rocks.

Some possibilities could be that your PI isn't generating a strong enough field, or the collapsing current di/dt is not sufficient, or the coil and circuitry are not meant to collapse and measure that quickly or perhaps the material was hematite and similar. I am not aware of any PI's coming from large corporations that allow and advertise their units can accurately work down to 1 usec. There seems to be a ton of smaller companies and some may actually allow such adjustments, but they probably would not work under 10 usec.
Paul

The equipment that I designed to operate down to a delay of 1uS is produced commercially as three widely different models by a well known company. The pulse transmitters are of low power but have a very high repetition rate. Sampling at 1uS is genuine

I witnessed the lag on an oscilloscope over one month ago. Who knows, maybe I have super magnetite. Paul

More likely is maghemite contamination that the geologist at the store was not aware of.

There could be a host of reasons why a certain PI unit doesn't report such lag from large magnetite. Other possibilities could be some GB cancellations or perhaps it's doing a good job at blocking magnetic materials.Paul

No GB or blocking in my tests.

Best method to really verify it is just get sufficient coil, generate sufficient field and quickly collapse it. Not difficult so far. Then view the output from a small coil on scope. View with and without the magnetite.Paul

Would this be by viewing an amplified signal from the coil? This I do, with both high power and low power transmitters and different pulse widths.

Also, make sure it's magnetite, not maghemite and not hematite. Big difference. Who knows, maybe I got ripped off and what I really have is maghemite, lol. Not sure except that I paid 30 bucks for it! Darn geologist! They're the greatest.Paul

My sample is a very pure magnetite-apatite mix, containing more than 60% iron and an average of 0.9% phosphorus.

Ahh, you're doing it wrong. You need to look at it exactly when the APPLIED magnetic field collapses. Two things occur when the current completely halts. 1. All remaining free magnetic fields collapse and convert into far fields, which collapses into the coil and continues to traverse through space. This takes not much longer than a few nanosec for say a three-foot coil. 2. The magnetic material lags caused by self-generating magnetic fields within the magnetic material itself. This simply depends on material type, size, and shape.Paul

This may well be happening, but if it is all over and done with before I sample, it doesn't matter. For any hobby detector application that I can think of, sampling at much less than 10uS would give rise to too many false responses and detecting frustratingly small scraps of metal.

:-) Now are you talking about natural magnetite or some commercial grade ferrite rod? Both contain lag, but commercial grades are usually light years better than natural stuff. Good point on your part anyways. Natural magnetite should vary. Just how much I'm not certain. Like I said, may be I have some rare super magnetite. Gosh, perhaps $30 was a bargain, lol.Paul

The rods and pot cores I use are obviously commercial grade ferrite. Natural magnetite is usually associated with maghemite and the proportions do vary. Soils and near surface material usually has a higher viscosity/susceptibility ratio than underlying geological material, due to oxidation.

Yes, but we're all interested in them since they're all natural and common. I know maghemite is common in good old California. Magnetite and Hematite should be common everywhere.Paul

They are.

Ferric Toes

[Paul]

Ferric Toes,

The geologist could indeed be wrong about the Magnetite he sold me, but odds are slim and none that he was wrong. In reference to your statement, "With that device I do see a good signal which decays into the noise inside of 10uS, but that is due to its conductivity, not viscosity." That statement indicates you PI is designed incorrectly. I've mentioned this to Carl in reference to the same problems that Rob & Reg are getting because I was surprised that people are getting such problems. Oh well.

Here's a very simple method of demonstrating the lag. Take a piece of that big magnetite and make a coil with it by simply wrapping wire around it. Then hook that up to your frequency generator with a small R and you will notice that at lower frequencies the magnetite inductor is mostly inductive, but at higher frequencies it becomes more resistive. You can tell this by the phase shift. Even the best loop sticks for radios begin to die out above 10 MHz. A 1 usec eddy decay contains appreciable 10+ MHz signals. Now a 1/2 inch diameter x 3" commercial grade loopstick is better than a hunk of natural magnetite.

Paul

[Ferric Toes]

That statement indicates you PI is designed incorrectly.
Paul

I am confident that the PI's I use are correctly designed and give the right answers.

Even the best loop sticks for radios begin to die out above 10 MHz. A 1 usec eddy decay contains appreciable 10+ MHz signals. Now a 1/2 inch diameter x 3" commercial grade loopstick is better than a hunk of natural magnetite.Paul

The ferrite rods (loopsticks) used in radios do not work in a PI. Although they work at radio frequencies, they give a very strong viscous signal on a PI. These are nickel/zinc ferrites. The type of ferrite that does work, is that which is used in switch mode power supplies and TV line scan transformers. This is called a "soft" ferrite because the domains can switch rapidly. and is made from a manganese/zinc material.

Ferric Toes

[Paul]

The ferrite rods (loopsticks) used in radios do not work in a PI.

I never said they do. I was comparing a loopstick to raw natural magnetite.

The following should put to rest the notion about raw natural magnetite. We were talking about large magnetite. Lets use a 7" wide chunk as example. You are actually telling me that 7" hunk of magnetite does not even lag less than 1 usec??? Lets say the lag is one tenth of a usec (100 nanosec). Lets look at this spectrum. We see the main frequency is 10 MHz followed by 20 MHz and so on. There are appreciable signals beyond 100 MHz. Even if the magnetite were only 90% efficient you would still see a lag. You are trying to tell me that a 7" hunk of raw material would make nearly a 100% efficient 100 MHz transformer. Sorry and no offense Terric Toes, but that is not going to happen anytime soon. In fact, it will not even come close to 100% efficiency at 1 MHz. 7" ???

Now, could it be that your PI possesses the same flaws that Regs and many others at this forum? Just food for thought.

One thing I am certain about, a 7" hunk of nature raw magnetite out in the desert will definitely contain appreciable lag at 1 MHz not to even mention 100 MHz!


Paul

[ElectroNovice]

Quote from paul
"Now, could it be that your PI possesses the same flaws that Regs and many others at this forum? Just food for thought."

Well Paul what are the design flaws in Erics or Regs designs?
How might they be corrected?
Is there a Pi in the market that does not have these flaws if so which one is it?
Do you have a design or suggestions or product where these flaws have been overcome?

If Erics design in the GS5 is so flawed then why does it perform so much better than a ML GP or SD over highly magnetic soil terrains?

Can you explain why Pi detectors can cope with large variations in many magnetically susceptible soils but run into problems with some soils that have much lower magnetic susceptibilities and much smaller lateral variations in magnetic susceptibility?

What TX frequency or Tx wave form (ie SQ wave .1/2 sine wave etc) do you think optimal for minimizing the effects of magnetically susceptible soils?

What Tx pulse length do you think is optimal optimal for producing the late time exponential decay of large nuggets say over 10 ozs? ie what is the optimal relationship between TC and Tx pulse length to produce a late time eddy current?

Have you measured the TC of any natural gold nuggets?
I could ask a heap more questions but I need to go to bed.
ElectroNovice

[robby_h]

You say I was having trouble! When? I never mentioned my solution or attack on the coil's change in properties once.
I said that I can virtually submerse the coil in a very large quantity of magnetite and see no magnetic lag at 10usec but you say this is not so???
You now say..."I witnessed the lag on an oscilloscope over one month ago. Who knows, maybe I have super magnetite"...
Or was it that you didn't take into account contamination, eddy currents or the obvious change you would have seen in the coil's inductance if you had say measured this with a meter? This also applies to your...
"Best method to really verify it is just get sufficient coil, generate sufficient field and quickly collapse it. Not difficult so far. Then view the output from a small coil on scope. View with and without the magnetite."
You haven't told us how you would know what you are actually looking at here or how you would split this observation into its true components.

It also depends on which point in the circuit you saw this lag. If what you saw was a physical widening of the spike itself then it wasn't magnetic lag at all. The rx signal is made up of components, some from the ground itself and some from the coil and you indicated you had a solution for the latter but we are yet to hear this.

Practically everyone here has done their share of reading re the "theory" on the excitation and decay of these domains and this is just going over old ground unless it leads to something not readily known.

Nothing occurs "instantly" but surely we can use that term relatively, especially if we have taken an effect, even long gone before sampling begins, into account?

Now, maybe you could offer something practical before this thread ends. Your "obvious" solution to our coil "problem" would be a nice start.