Fast coils: worth it?

[bklein]

I wish I knew exactly how this works.
It is a proven fact that "fast" coils are needed for detecting targets that have a TC of less than 10us.
I have also noticed that if I reduce the bandwidth of the preamp below 150KHz, the signal of targets that have a fast decay, is attenuated.

There are other things involved, like the PPS or pulse per second (maybe 3000) and the sweep speed.
A standard sweep speed of 1m/s, amounts to about 3Hz of signal response for a 300mm diameter coil.

We usually sample at the rate of the PPS and then integrate the samples.

Monolith

I would say that we are most interested in obtaining the fast response and as such that would correspond to higher frequency. Swinging the coil and the pulse frequency itself are more related to the detection and audio circuits, not the coil. Make believe the target is a 75khz oscillator out there in the dirt for example (really it only "transmits" after experiencing the transmt excitation pulse). We don't want the 60hz power lines above us to overwhelm the 75khz oscillator signal. I think we want a coil optimized for whatever the signal spectrum is centered upon. A resonant coil would improve reception but may not be the right design for transmit. Just a lot of things to consider, which is why I find this area of electronics fascinating. I borrowed a spectrum analyzer from work and would like to characterize the eddy current signal from a gold ring vs a nugget. Anyone have any ideas on how to probe it? Anyone done this before?

[WM6]

. I borrowed a spectrum analyzer from work and would like to characterize the eddy current signal from a gold ring vs a nugget. Anyone have any ideas on how to probe it?

Very easy. You need MD receiver stage (as you have it) without any filter (if exist you need to bypass it). Then connect your spectrum analyser at output of first amplifier.

As first, scan spectrum without any target to get spectrum difference to comparition your useful signal with pre-existing signal. To catch harmonics you (your sp. analyser) have to sweep at least between 100Hz and 20kHz.

[Monolith]

I would say that we are most interested in obtaining the fast response and as such that would correspond to higher frequency. Swinging the coil and the pulse frequency itself are more related to the detection and audio circuits, not the coil. Make believe the target is a 75khz oscillator out there in the dirt for example (really it only "transmits" after experiencing the transmt excitation pulse). We don't want the 60hz power lines above us to overwhelm the 75khz oscillator signal. I think we want a coil optimized for whatever the signal spectrum is centered upon. A resonant coil would improve reception but may not be the right design for transmit. Just a lot of things to consider, which is why I find this area of electronics fascinating. I borrowed a spectrum analyzer from work and would like to characterize the eddy current signal from a gold ring vs a nugget. Anyone have any ideas on how to probe it? Anyone done this before?

After you switch off the TX pulse, the coil carries the decay of the Flyback. This decay is exponential, but it will not follow a smooth curve at the output of the preamp, where we look at it, because it is influenced by many factors, among them:
The coil self resonant frequency.
The damping resistor.
The input protection diodes that stop conducting at about 0.7V.
The saturation of the preamp.
Cable and wire stray inductances and capacitances, they cause oscillations.
The response of the soil.
The response of the target.

Each shape and volume of target, even of the same metal, gives a different response.

When you talk of the same metal this is very strict. For example: the US$ 5c coin and the US$ 10c coin are of nearly the same alloy, only about 3% difference, but they give a very different target signature.

Or, take a 18 Carat gold ring. You can get 18 Carat gold in white gold, pink gold, yellow gold shades, to name only a few. They are all genuine 18 Carat, but each of these shades of gold, and there are about 100 varieties, even with the exact same ring dimensions, will give a different signature.
To complicate matters a bit more, the position of the ring, flat, at 90 degrees or any angle between will again change the signal response dramatically.

In general the most important factor is the decay time of the target signal that is produced by the eddy currents in the target.

The decay time of the US$ 5c coin mentioned above, is about 15us. The decay time of the US$ 10c coin is about 100us, in spite of being very much smaller.

So, unless you design the detector for one very specific target, all detectors need to be designed to be able to detect targets from as small as possible to fairly large.
We could say with a TC from 5us to 500us would be a common range.

Monolith

[Noemio]

back to the original request

As I wrote at the beginning of this 3D, I was looking for “objective data” and possibly “comparative data”. And more, if it’s possible, I would like to know which MD are suitable to be coupled with such fast coils.

Last years I built a Gary’s PI MD, a Stuart II (without great success), a Hammerhead, a couple of Delta Pulse and last (but not least) a Surfmaster (I liked Blind Squirrel and Skif-PI too; maybe in the future…).

For my Delta Pulse I wanted to try a fast coil, and so I:

- read several times bbsailor’s article
- compared dielectric constant of various materials to find the best and cheapest choice
- bought twisted pair
- bought polyethylene spiral wrap
- bought 3M Scoth 24 for the shield
- compared capacitance values of various coax cables and chose RG59 (which has less than 70pF/m vs about 100, or more, of RG58!)
- modified "a bit" frequency a minimum delay of my MD

and finally built my “fast coil”.

Apparently I don’t notice any particular difference. I reached the value of 15 µs with both the coils: the classic coil built with just enameled wire and no shield and the fast coil built with all the necessary features.

At this point I should have done all kinds of accurate measurements and wonder about a reason of this… but I’m a bit lazy so I decided to open this 3D to know if anybody already did it.

I'm glad to read it again a lot of theory about fast coils, but I’ll be happy to read somebody talking about more practical results.

Thank you in advance


p.s.: teflon has a dielectric constant of 2.1, polyethylene has 2.3 but Kynar has 6.4 or even more!!!

[bklein]

wow, you've built a lot of stuff!
So... what if you then try using 2 coils. One is the transmit coil and the other (test) coil is the "fast" one or faster/slower one. Bring the sample timing signals over to the RX coil circuit. Then maybe you can determine if one coil is really better than another? Never heard of someone doing this but.... kind of like running 2 detectors in parallel.

Like I was getting at (maybe I should start a separate thread): the coil has a transmit function and a receive function. I'd like to know just what type of signal or coil sending it will best stimulate the gold target. Then I want to know what coil and associated circuitry will best detect it over the damn nails and hairpins I'm dealing with now.

I'm thinking I should be able to hit the target with the normal Surfmaster tx signal, bring the timing signals out to an identical rx dedicated circuit but place the rx coil at a 90 degree angle to the the target and tx signal. Would this help isolate the rx coil from the tx signal so I can focus on the rx signal?

Seems to me we are trying to get this dual use (tx/rx) out of the coil and the tx causes the HV that is a pain in the butt to deal with. A separate rx coil doesn't have the HV pulse so you can concentrate on optimizing it for rx and later do optimization on a coil for tx....

Barry

[Tinkerer]

As I wrote at the beginning of this 3D, I was looking for “objective data” and possibly “comparative data”. And more, if it’s possible, I would like to know which MD are suitable to be coupled with such fast coils.

Last years I built a Gary’s PI MD, a Stuart II (without great success), a Hammerhead, a couple of Delta Pulse and last (but not least) a Surfmaster (I liked Blind Squirrel and Skif-PI too; maybe in the future…).

For my Delta Pulse I wanted to try a fast coil, and so I:

- read several times bbsailor’s article
- compared dielectric constant of various materials to find the best and cheapest choice
- bought twisted pair
- bought polyethylene spiral wrap
- bought 3M Scoth 24 for the shield
- compared capacitance values of various coax cables and chose RG59 (which has less than 70pF/m vs about 100, or more, of RG58!)
- modified "a bit" frequency a minimum delay of my MD

and finally built my “fast coil”.

Apparently I don’t notice any particular difference. I reached the value of 15 µs with both the coils: the classic coil built with just enameled wire and no shield and the fast coil built with all the necessary features.

At this point I should have done all kinds of accurate measurements and wonder about a reason of this… but I’m a bit lazy so I decided to open this 3D to know if anybody already did it.

I'm glad to read it again a lot of theory about fast coils, but I’ll be happy to read somebody talking about more practical results.

Thank you in advance


p.s.: teflon has a dielectric constant of 2.1, polyethylene has 2.3 but Kynar has 6.4 or even more!!!

Noemio,

You have built many detectors, but I don't know what your specific goal is. What are you searching? Gold nuggets? Coins? Deep buried large targets? Gold and silver chains? Lost jewelry?
Fast coils are for small targets. But it is not the coil alone that makes a detector good for small targets.
The circuit needs to be optimized for small targets too.
The TX has to be switching fast enough.
It needs high PPS.
A combination of low power and high PPS usually works best.
The preamp needs to be fast too, if you have a preamp that stays in saturation for several microseconds, it is no good for small targets.

So, maybe you want to give us some details of what your goals are, then we might be able to help achieving these goals.

Tinkerer

[Tinkerer]

wow, you've built a lot of stuff!
So... what if you then try using 2 coils. One is the transmit coil and the other (test) coil is the "fast" one or faster/slower one. Bring the sample timing signals over to the RX coil circuit. Then maybe you can determine if one coil is really better than another? Never heard of someone doing this but.... kind of like running 2 detectors in parallel.

Like I was getting at (maybe I should start a separate thread): the coil has a transmit function and a receive function. I'd like to know just what type of signal or coil sending it will best stimulate the gold target. Then I want to know what coil and associated circuitry will best detect it over the damn nails and hairpins I'm dealing with now.

I'm thinking I should be able to hit the target with the normal Surfmaster tx signal, bring the timing signals out to an identical rx dedicated circuit but place the rx coil at a 90 degree angle to the the target and tx signal. Would this help isolate the rx coil from the tx signal so I can focus on the rx signal?

Seems to me we are trying to get this dual use (tx/rx) out of the coil and the tx causes the HV that is a pain in the butt to deal with. A separate rx coil doesn't have the HV pulse so you can concentrate on optimizing it for rx and later do optimization on a coil for tx....

Barry

Hi Barry,

Separate RX coils have advantages, but you need to decide which configuration you want to use. Induction Balanced or not.

With the induction balance you get very little residual Flyback in the RX.

If the coils are not induction balanced, the decay of the Flyback is nearly the same in the RX coil as in the TX coil, due to the coupling factor (k).

If the TX and RX coils have a very different TC, strange interferences happen that are sometimes difficult to unravel.

For small gold nuggets you want a very fast detector. Very fast means sampling very soon, maybe 5us after TX switch off. It also means quick response. You want to hear the audio indicating the nugget while the coil is still above the nugget, not 500ms later.

So tell me a bit more of your circuit and I might be able to tell you how to adapt it for your purpose.

Tinkerer

[bernte_one]

Interesting thread...fast coils. Have builders noticed better performance using spiral type coils or other variations that reduce wire-wire capacitance? The concept is not new...a half century ago hams and antenna engineers created magnetic loop antennas (Google this phrase) in this manner using either face wound (spiral) or edge wound conductors. By carefully spacing the wires, the null (and node) could be clearly defined. The same goes for shielded loops. I'm not certain there is a relationship to the MD coil, but suspect there is.

I have built several loop antennas over the years...now about to make some PI coils. I look forward to playing with these as well as my first MD (a Surf PI)! The trick is not to let the smoke out.

Peter

i have build some flat planar coils the performance i get seems to be equal
somes builder on german pulsdetekor forum swear of spiral shaped (planar and co planar coils) they have compared it with other they have biuld and got more sensibility the magnetic pulse field is more uniformly

http://www.pulsdetektor.de/apboard/b...id=13&BoardID=

http://fotoalbum.web.de/gast/normanbsupski

http://www.metdet.ru/Sensor_K1.htm

[Ism]

As I wrote at the beginning of this 3D, I was looking for
I'm glad to read it again a lot of theory about fast coils, but I’ll be happy to read somebody talking about more practical results.

Thank you in advance

In practicality (for mono coils), a faster coil allows for faster damping. Faster damping allows for better sampling of the signal. If its a small low conductor, the signal decays too fast for a slow coil to settle down enough to see it after the pulse. So for those searching for gold, its very important.

The faster coil also allows the detector to see the signal at a greater amplitude as it is decaying. I believe this equates to better depth. But remember, a fast coil is only as good as the instrument its attached to. I hope this is a practical explaination.

[bklein]

Hi Barry,

Separate RX coils have advantages, but you need to decide which configuration you want to use. Induction Balanced or not.

With the induction balance you get very little residual Flyback in the RX.

If the coils are not induction balanced, the decay of the Flyback is nearly the same in the RX coil as in the TX coil, due to the coupling factor (k).

If the TX and RX coils have a very different TC, strange interferences happen that are sometimes difficult to unravel.

For small gold nuggets you want a very fast detector. Very fast means sampling very soon, maybe 5us after TX switch off. It also means quick response. You want to hear the audio indicating the nugget while the coil is still above the nugget, not 500ms later.

So tell me a bit more of your circuit and I might be able to tell you how to adapt it for your purpose.

Tinkerer

What I am suggesting is that we separate the TX and RX functions for purposes of optimization. I don't know - maybe not worth the trouble - but I really would approach this looking at each function separately. Learn what coil characteristics work best for each function. Document what those capabilities are. So in this light, the tx coil and rx coil are to be isolated as much as possible so that the tx pulse does not get imposed on the rx one. Thus my suggestion (by a newbie) about placing them at 90' angles. Maybe even just magnetically shield the rx coil from the tx signal in some way. Just get things dialed in for each function then you know the best you can do for each mode.

[Tinkerer]

What I am suggesting is that we separate the TX and RX functions for purposes of optimization. I don't know - maybe not worth the trouble - but I really would approach this looking at each function separately. Learn what coil characteristics work best for each function. Document what those capabilities are. So in this light, the tx coil and rx coil are to be isolated as much as possible so that the tx pulse does not get imposed on the rx one. Thus my suggestion (by a newbie) about placing them at 90' angles. Maybe even just magnetically shield the rx coil from the tx signal in some way. Just get things dialed in for each function then you know the best you can do for each mode.

The RX coil at 90 degrees, will receive very little TX field, but also very little target field.
2 box detectors use this coil arrangement, but I think in the opposite way.

Tinkerer

[bklein]

The RX coil at 90 degrees, will receive very little TX field, but also very little target field.
2 box detectors use this coil arrangement, but I think in the opposite way.

Tinkerer

I am not suggesting a complete detector solution here - I am suggesting a way to optimize the tx or rx design in a test bench environment. So, given that, I would think that the target response signal would radiate in all directions. In fact I'd would even like to validate that given the physical character of some targets....

Barry

[Tinkerer]

I am not suggesting a complete detector solution here - I am suggesting a way to optimize the tx or rx design in a test bench environment. So, given that, I would think that the target response signal would radiate in all directions. In fact I'd would even like to validate that given the physical character of some targets....

Barry

OK, for bench testing, build 2 identical coils, then place them at 90 degrees, 30cm from the target, so that the target is at the same distance from both coils.

Start with approximately spheric targets.

Then change to disks and then to rings.

You will notice big differences.

Then try different metals.

Observe the results.

Then make a small change on one of the coils and again observe the results.

Keep repeating until you have tried the extremes on either coil. Like 200 to 400uH and different wires, like Litz and solid.

For these tests do not shield the coils, just wrap them with spiral wrap so you can easily make changes.

There will be noise, that can be taken care later.

You seem to have the will power to go through the test runs. The results are going to be interesting.

Good luck

Tinkerer

Ah, if you post interesting results, you will get help from many forum members.

[bklein]

Exactly! I can't promise immediate results but we think alike - exactly the approaches I have been intending to do.

[Noemio]

last effort

First of all this could be the “hunting” area: http://maps.google.it/maps?hl=it&ie=...,0.703125&z=11

For sure Roman settlements and beaches at the end of bathing season.
This means that my targets could be coins or jewelry and metals like gold or bronze…
Not at all gold nuggets, never heard!

On bbsailor article I read “…can be used to locate coinsize objects, gold jewelry and gold nuggets.” This made me believe that, without this kind of “fast coils”, it would have been very difficult find gold objects.

Now, I know that my English is absolutely horrible (on the other hand I'm Italian...) and thus can be hard to understand what I mean trough my tarzan style writing, but what I would like to read is somebody who can make statements like “yes, first I built a coil just with enameled wire and wasn’t able to find, later I built a fast coil and now I can detect gold rings and jewelry and nuggets…. I use it on my xxxx homemade detector”. And, if possible, some other comparative data.

That's all. Easy, huh?