PI signal snapshots

[Unregistered]

The sampling problem of not being able to sample less than 10 usec has to do with a combination of things, mainly the coil. However, the amplifier used as a preamp also has a large part as does the FET and all associated circuitry. It is also extremely difficult to get a coil to dampen correctly in the time needed to samle less than 10 usec, simply because of the associated capacitances, among other things.

I see. Hypothetically, lets say someone goes all out and spends a fortune to achieve a really clean fast pre op-amp that can sample at say 1 usec in addition to a small coil that allows such quick sampling. Could we run into another problem? That problem being hot-rock signals. As you know, hot rocks are magnetic materials, or at least by my definition. Sure, a perfect ideal hot rock would not generate any signal after the coils magnetic field collapses, but in real life hot rocks aren't so ideal in that they can only collapse magnetically so fast. Hence, if you magnetize a hot rock and then instantly remove that magnetizing field, then it takes a certain amount of time for the hot rock to complete decay. I'm wondering just how long it takes. Question: have you seen any scope pics of hot rocks or heard mention of any data regarding how long it takes the hot rock signal to appreciably decay? Key word is "appreciably." If you have a really small nugget that decays at a high rate, then hopefully the hot rock signal has decayed appreciably enough so the detector can still pick up the small nuggets eddy decay. In that sense, do you have a feel for about how quick a PI can sample before it appreciably picks up on the hot rock signal. This is a relative question. Lets just base it on the smallest nuggets that VLF's can detect. I'm asking because I have no idea. For all I know it could be in the nano second range.

Thanks,
Paul

[Reg]

Hi Paul,

You are correct, most hotrocks, are not comprised of the same material and some will respond differently than others. Now, with that being said, a typical hotrock, one with a lot of magnetite will generate little or no response even down to the 5 usec level on a PI, providing one takes a second sample to eliminate the earth's magnetic field effect. This same rock would cause an extreme response on a VLF if the ground balance were off just a little.

On a PI, any residual magnetic response within the rock would be canceled by the second sample, so that wouldn't be a problem.

Unfortunately, some of the hotrocks that are primarily magnetite have other components in them and those items may readily cause a response, so there will be some rocks that will still be a problem.

Fortunately in most places such rocks are quite rare. As an example, one location, Gold Basin has a lot of "hotrocks" that give VLF's fits. It is easy to come upon large areas of them.

Now, one can hunt these places with ease with many PI's, simply because the hotrocks are ignored. One might run into 3 or 4 total hotrocks that generate a strong signal during the day.

Normally, if a rock causes a response on a PI, it is either a form of basalt, or it is a meteorite.

Getting back to sampling at extremely short delays, it is extremely difficult to get the decay curve to decay fast enough to sample at less than 5 usec. Actually, it extremely difficult to get below 10 usec, so obtaining a 5 usec response is more of an art. For those interested, bbsailor has written quite a bit about it on this forum.

Personally, I see little need to sample at less than 5 usec. This delay will allow the detection of gold of less than a grain in size. Any gold requiring a shorter sample will take longer to find that it is worth. This type of gold would be better found using a dry washer or dredging.

Reg

[ma330]

similar ic

hi to all:
i build bandido II umax.but not found ic(mc33178-2931a-tlc2262)
please send for me similar ic for (mc33178_2931a_tlc2262)
tank you.good by

[Paul]

Thanks for the info Rec.

Does anyone know where the sample delay adjust is on a Minelab GP3500 or GP3000? I'm not sure, but the closest thing seems to be the Iron Discriminate dial.

Paul

[Paul]

Reg,

This seems like a question up your alley. I always thought magnetite was more magnetic than hematite. I have big chunks of both. The hematite is hardly attracted to a permanent magnet, but the magnetite is strong almost like pure iron. Then there's a third type, maghemite. As you know, if these materials were perfect then there would be zero delay after to current pulse, but in real life they're not. I read that magnetite hardly, if at all, affects a PI ... hematite somewhat affects a PI ... and maghemite can drastically affect PI's. Hmmm, now I'm confused because I thought magnetite had the strongest magnetic properties and therefore would have thought magnetite would have the longest magnetic-lag / signal-delay. I understand these lags / delays vary with material size, but can you offer some typical delay time of these three materials.

Many thanks,
Paul

[robby_h]

Magnetite is classed as having a very high susceptability to a magnetic field but it's decay is almost instantaneous. The problem though is that it changes the coil's inductance much more than that of ironstone ground with just a long decay component although the two usually occur together. Any large change in inductance is relected in the spike's decay time and so adds another component that must be dealt with in the Minelab design where early sampling is desired and this problem is compounded even further if using different pulse lengths. The designs that use single pulse length, constant current tx or flat top pulses suffer less.
If you sample the curve to zero instead of waiting until it levels then you would get Reg's "hypersenitive ground effect" on magnetite or its cousins as you are then sampling the coil's change in inductance while also adding this to the ground's decay and this effect can still occur even if you wait until after the signal levels off in some cases. This effect can't be dealt with by the GB circuit as it is gone before the later ground sample is taken and even if it wasn't, it would not form the same proportions.
Rob.

[Paul]

I see. So basically you're saying the unit is getting in the way of it self. That makes sense. What a relief! I was wondering how in the world maghemite could decay hundreds times longer than magnetite or hematite. I thought perhaps maghemite contained some pure iron that was electrically conductive. As I understand, iron decays like 10+ times longer than natural gold. I think the reason is two fold. 1. Pure iron is a better conductor of electricity than natural gold. 2. Pure iron is magnetic, and not only is it magnetic, but it can have extremely high permeability. In fact, really pure iron has over 200,000 permeability. Of course, to achieve such high "effective" permeability you would need an extremely long iron rod. So a long piece of pure iron could realistically have an effective permeability in the hundreds to perhaps a thousand or more. That equates to a lot of inductance.

What about highly mineralized wet soils that are fairly decent conductors of electricity? I'm thinking the resistance would still be relatively high, but the ground is 1000's times larger than a small nugget. Could an eddy in such ground generate a long decaying significant signal or is such soil a rarity? Just trying to think of all worst-case scenarios before designing a metal detector.

One final issue is dielectric material such as water. As you know dielectric materials causes an opposite effect than magnetic materials. Even water possesses a lag / delay, but I have no idea to what degree. I'm wondering if large amount of water could generate a significant decay signal.

Thanks,
Paul

[robby_h]

I think we are getting our wires crossed a bit here. I only mentioned magnetite and maghemite is a different kettle of fish as it doesn't decay at pulse switch off.
Most people get hold of and study some hot ground samples which are a mixture of the minerals and the effects that need cancelling. Ironstone soil is always at least mildly conducting but usually decays before a sample is taken. Wet salt lakes do cause problems for a pi and the only fix to date is to use so called salt coils which are sometimes figure 8. Wet sea side sand and sea water also gives problems if sampling early.
The problems involved show up when you attempt a design as you will find out for yourself.
I found a piccy that I put together some time ago and re-edited the layers slightly for this thread.
It is basically self explanitory I hope. This design uses inductance limiting tx similar to Ml's so there is a huge amount of energy in the long pulse's spike compared to that of the short pulse as the current at the end of the pulse climbs to vt/l. The tx methods that use resistance limiting may show much less difference in spike width when using different pulse lengths because the rise to maximum occurs in the same time.

[Reg]

Hi Paul,

I think you are getting magnetic properties mixed up with properties that affect the decay of the signal. In the case of ground signals or maybe even rock signals, the term commonly used to refer to its altering of the decay signal is its magnetic viscosity. In basic terms, this magnetic viscosity could be also stated as its target response, whereby any decay change caused by any non metal object or metal object is classified as a target response.

As Robby said, magnetite may have strong magnetic attraction and may alter the magnetic field, it still doesn't present much of a problem to a PI if the sample is taken on the flat line where the decay has come back up and leveled off. Again, the reason is the magnetite signal comes and goes almost instanteously, so its target response is long gone when the sample is taken. As such, no signal occurs.

Now, on a VLF where all signals are evaluated and there is no delay between transmit and recieve, but rather a phase shift, then the magnetite properties will clearly display themselves as strong signals. Thus, magnetite can cause a very strong signal.

Getting back to a PI, I am not sure just how fast the magnetite decays. I don't have the capability of sampling fast enough to see it. However, as mentioned, if you sample on the decay curve, then magnetite will or can alter that sufficiently to cause problems.

Now, referring to Robby's pics, they display a couple of things that one should realize. The items are not that obvious, but are there. The difference between pic 1 and 2 is minimal, but does display a slight difference in rise time of the signal out of the preamp. This shows that the shorter pulse could be sampled a little bit sooner than the long pulse. The time difference is probably only a usec or two, though.

Now pics 4 and 5 clearly show the advantage of reducing the gain of the preamp. The signal clearly rises faster which would allow for a shorter time before sampling if necessary. This normally occurs because the gain characteristics of the opamps involved are severely limited, speed wise. So, when comparing the sets of pics, one can see the distinct advantage of reducing the gain out of the preamp as a primary means of allowing for a faster sampling time. It also shows that simply changing the gain will also change the damping of the coil as seen at the preamp output. If you look carefully at the signals, you will see a small overshoot before the signal levels off. This indicates the coil is very slightly underdamped. When looking at the two pics, 4 and 5, one can see that the longer pulse now causes a greater overshoot (at the point where it levels off).

What isn't displayed is the fact that as the gain is reduced, the bandpass is usually widened. Yes, it does show that the IC comes out of saturation faster, but the gain may be different during the first few usec's on a very high gain circuit. This can cause some strange results. Eric Foster posted some results along this line quite some time ago.

Now, where am I going with all of this? Well, what I am trying to point out is the answers you may be looking for are not as simple as just saying this rock is at fault, or that is at fault, or this should happen or that should happen. There are so many interactions that it is scary. This makes it very difficult to determine just what is happening or what is the best ultimate solution.

Reg

[Paul]

Thanks for pics.

I think we are getting our wires crossed a bit here. I only mentioned magnetite and maghemite is a different kettle of fish as it doesn't decay at pulse switch off.

Actually that's not true. If you take a full spectrum of magnetic material one learns that all magnetic materials begin to phase from reactance to resistance. This is due to magnetic lag, a real phenomenon. This lag is seen just after the current pulse ends. Just how much it lags greatly depends on how fast the current pulse decayed and the size, shape of magnetic material and type of material.

I think you are getting magnetic properties mixed up with properties that affect the decay of the signal. In the case of ground signals or maybe even rock signals, the term commonly used to refer to its altering of the decay signal is its magnetic viscosity. In basic terms, this magnetic viscosity could be also stated as its target response, whereby any decay change caused by any non metal object or metal object is classified as a target response.

Reg, I appreciate your replies, but don't really appreciate being called "mixed up," especially when you confirm what I said. Anyhow, I've always known it as "Magnetic Lag."

As Robby said, magnetite may have strong magnetic attraction and may alter the magnetic field, it still doesn't present much of a problem to a PI if the sample is taken on the flat line where the decay has come back up and leveled off.

Wow, actually Rob said "I only mentioned magnetite and maghemite is a different kettle of fish as it doesn't decay at pulse switch off." Are you also suggesting there's no magnetic lag in magnetic materials?

Again, the reason is the magnetite signal comes and goes almost instanteously

"Almost?" Why call me "mixed up" when you just agreed with what I said and completely disagreed with Rob? My question was, "I understand these lags / delays vary with material size, but can you offer some typical delay time of these three materials." That was in reference to magnetite, hematite, and maghemite.

Now, on a VLF where all signals are evaluated and there is no delay between transmit and recieve, but rather a phase shift, then the magnetite properties will clearly display themselves as strong signals. Thus, magnetite can cause a very strong signal.

It has everything to do with magnetic lag. Higher frequencies equates to more magnetic lag, which equates to resistance. A magnetic core will heat up if you crank the frequency up in a strong signal even if the core isn't touching any current carrying wires.

Getting back to a PI, I am not sure just how fast the magnetite decays. I don't have the capability of sampling fast enough to see it. However, as mentioned, if you sample on the decay curve, then magnetite will or can alter that sufficiently to cause problems.

You don't think a huge chuck of magnetite will lag beyond 10 usec?

Thanks,
Paul

[Reg]

Paul,

I didn't call you mixed, but take it any way you like. Obviously, you have all the answers so why ask the questions? Feel free to apply your knowledge and build the detector of your choice. I see little need to continue this discussion.

Reg

[Paul]

I didn't call you mixed, but take it any way you like.

So now you're calling me liar, lol. Here's your quote, the first line of your previous post:

I think you are getting magnetic properties mixed up with properties that affect the decay of the signal.

Ask any English major what it means. Look up what "mixed up" means. It's when a person gets things mixed up. Your statement is very clear.
Funny how you insult me (twice), but you agreed with me. Yet you disagreed with Rob.

Obviously, you have all the answers so why ask the questions? Feel free to apply your knowledge and build the detector of your choice. I see little need to continue this discussion.

Wow, such fuzzy logic. My questions are relating to values / data, not theory design. I'm interested in data, but if you want duel tech talk then lets go at it guy.

Paul

[Unregistered]

Getting back to a PI, I am not sure just how fast the magnetite decays. I don't have the capability of sampling fast enough to see it.

Thank You! That's all you had to say. I'm dumbfounded why you found the need to insult me by stating I'm mixed up when you clearly demonstrated we were in agreement.

Please check yourself Sir.

Paul

[Carl-NC]

Paul, I don't see where Reg's very detailed and time-consuming replies have justified your reaction. No one has called you any names, so please settle down.

Thanks...

[Reg]

Hey Robby,

I like your pics. They tell me a lot and also want me to ask a couple of questions. What opamp was used in the pics and the second question was how much was the gain reduced on the pics 4 and 5? The 50 usec pulse pic shows a signficant reduction in recovery time.

Cheers,

Reg