detection depth

[Skippy]

Re: the foil. I found 3 different samples of 'regular' grade thin foil. They measure up as 0.0103mm ; 0.0109mm; 0.0138mm. I don't have anything to hand that would be sold as 'thick', 'strong', 'turkey roasting' etc, though I will have a look tomorrow. I've also managed to bag some thick foil sheet from a friend who works in the catering industry. It's from a 'silver' serving platter as used at wedding buffets etc. And it's a substantial 0.165mm thick. So I've got a pretty decent spread of thicknesses now. It will be interesting to see how good a straight-line they give for TC vs. thickness. The info on the web suggests it's not all 1000-series alloy (ie. near-pure alloy), so resistivity could be noticeably different.

Re: your previous decay/TC charts: Is there some labelling missing from the 'top right' one?
This tempco effect isn't helpful, I do my tests in a moderate indoor environment, and I've not considered cold or hot tests.

I was thinking about the non-linear decay's you're seeing. It seems logical to me that whether a thin sheet is 'thin enough' will depend on the linear size of the sheet. So for example a 25mm square sample may give a straight-line decay with a thickness up to 1mm, whereas a 10mm square may only be linear up to 0.4mm thickness. Beyond some certain thickness/size ratio, you'll start to see 'bulk' effects, and non-linearity.

[Skippy]

Gotta say, we seem to be drifting off-topic a bit, discussing piddly bits of ally foil in a thread about finding things deep with big coils.

I was intending to post my general target tests in this thread:
http://www.geotech1.com/forums/showt...6-test-targets

[green]

Gotta say, we seem to be drifting off-topic a bit, discussing piddly bits of ally foil in a thread about finding things deep with big coils.

I was intending to post my general target tests in this thread:
http://www.geotech1.com/forums/showt...6-test-targets

I agree about the drifting part but the thread title is detection depth. Could be about increasing the detection depth of a 1grain nugget at 2inches or a US quarter at 13inches starting with a 8inch coil.

I would like to know the best way to change my detector to detect a quarter at 24inches, another forum suggested it was possible with a GPX4500 with an 11inch mono coil.

Including the HyperPhysics calculated chart and a chart with my detector from awhile back. From my detector chart, US quarter at 13.5 inches. HyperPhsics chart, 13.5 inches on the 8inch coil gives a signal amplitude of about 6. Think that defines the noise level so I need an amplitude of 6 at 24inches. I'm using two 8inch round coils, figure8 Rx to reduce noise. Tx_1000pps,160usec on time, 6250A/sec for 1A peak, 80ma average coil current. The coil charts the same as a 8inch mono with HyperPhysics. If I wound a 11 inch coil, same inductance I would get about .5 for the signal at 24inches. Would need to increase S/N by 6/.5(12 times). Any thoughts on the best way to get there? Or maybe I'm looking at it all wrong and someone could straight me out.

[Skippy]

"Any thoughts on the best way to get there?"
That's a question for the PI experts, I can't offer any useful advice. Don't the GP machines use high coil currents?

[green]

"Any thoughts on the best way to get there?"
That's a question for the PI experts, I can't offer any useful advice. Don't the GP machines use high coil currents?

Thanks for the reply. 12 times the current doesn't seem practical so I keep wondering if I'm missing something simple. Still hoping someone can help me.

[Chiv]

Looking at your chart Green, I was wondering that crossover point on 300mm to 500mm coils inductance about 190/200mm is some type of sweet spot, if you could send that sweet point cross over point high in depth with changing the inductance would you get deeper with your targets, as graph with same turns with coil size increases inductance. I watched these blokes vid ( https://detectormods.com) on some of the depth they get with modding their gpx and it seem that they concentrate more on the internals more than coil.

[green]

Looking at your chart Green, I was wondering that crossover point on 300mm to 500mm coils inductance about 190/200mm is some type of sweet spot, if you could send that sweet point cross over point high in depth with changing the inductance would you get deeper with your targets, as graph with same turns with coil size increases inductance. I watched these blokes vid ( https://detectormods.com) on some of the depth they get with modding their gpx and it seem that they concentrate more on the internals more than coil.

The crossover point, distance where two different size coils give the same return signal(same inductance, same current profile) is about equal to (coil diameter A + coil diameter B)/4. (300mm +500mm)/4=200mm. I would expect a 1meter coil and a 500mm coil to cross at (1000mm + 500mm)/4=375mm.

[you will be well worked digging everything...especially iron...way more junk than most people realize...18 to 24 inches on quarters is not unreasonable...take a good shovel!!!] statement from another forum discussing GPX4500. Searched GPX4500, has a 11inch mono. 18 to 24 inches is a fair spread and quarters is plural so I don't know if it could detect one quarter at 24inches.
Might have assumed wrong, looked up the 4500 again and noticed a 18inch mono coil as being an option. Maybe the 11inch coil can't detect the quarter at 24inches. The detector has a 68AH battery so I would think the Tx current is higher than my detector, still probably not high enough to get my detector to detect a quarter at 24inches with a 11inch coil. Would be interested what the detection distance in air is for a US nickel, US quarter and 1x1inch piece cut from an aluminum drink can if someone has a GPX4500(any or all of the three targets). I keep thinking signal strength should be physics. Detection distance includes S/N ratio. Maybe lowering noise is where some magic happens.

[green]

Try different Tx on times and shape with two different coils, (1000pps, 6250A/sec, 160usec on) (10pps, 62.5A/sec, 16000usec on) (10pps, 1A constant current, 16000usec on).

from reply#71

A test comparing (1000pps, 6250A/sec, 160usec on)Tx and (100pps, 625A/sec, 1600usec on)Tx with a US quarter and 1oz copper coin. Target was centered on one of the figure8 Rx coils(same location for each test). Times were close, not exact. Constant current peak 1/2 constant rate peak(same average coil current). The longer Tx pulse did increase the return signal. For the same average coil current cr is better. Thinking of ordering a 1inch copper cube and 1.5inch copper cube to see how much better 16000usec Tx would be with longer TC targets. Don't know if 10pps would be useful?

[green]

Try different Tx on times and shape with two different coils, (1000pps, 6250A/sec, 160usec on) (10pps, 62.5A/sec, 16000usec on) (10pps, 1A constant current, 16000usec on).

from reply#71

A test comparing (1000pps, 6250A/sec, 160usec on)Tx and (100pps, 625A/sec, 1600usec on)Tx with a US quarter and 1oz copper coin. Target was centered on one of the figure8 Rx coils(same location for each test). Times were close, not exact. Constant current peak 1/2 constant rate peak(same average coil current). The longer Tx pulse did increase the return signal. For the same average coil current cr is better. Thinking of ordering a 1inch copper cube and 1.5inch copper cube to see how much better 16000usec Tx would be with longer TC targets. Don't know if 10pps would be useful?

Tx on time and shape, constant rate or constant current effect target Rx signal amplitude. Are there other waveforms that might give a higher Rx signal for the same average coil current? Average coil current=average on current*%on time. What would be the best %on time? Any thing to consider before the preamp besides coil diameter and Tx waveform?

[Ferric Toes]

Tx on time and shape, constant rate or constant current effect target Rx signal amplitude. Are there other waveforms that might give a higher Rx signal for the same average coil current? Average coil current=average on current*%on time. What would be the best %on time? Any thing to consider before the preamp besides coil diameter and Tx waveform?

There is a lot of good information on the internet dealing with 'transient electromagnetic ore body detection' This is going to the other extreme, but the technology is basically the same, but much scaled down in our case. In fact, geophysics is also moving into the 'near surface' applications.

Have a look at the following - Google 'Broadband Transient Electromagnetic Techniques'. Scroll down to the .pdf by Zonge where it talks about 'Improved Methods for Detection and Classification of UXO's'. The merits of the half sine pulse method are discussed and how this results in a greater transmitted moment for the same current draw. You will need to magnify the on-screen text and pictures to see the considerable detail on the poster.

Another paper that is interesting and deals with an unusual coil arrangement is - 'Opposing -coils transient electromagnetic method focused near-surface resolution'. Again, this is a downloadable .pdf.

Eric.

[green]

Tx on time and shape, constant rate or constant current effect target Rx signal amplitude. Are there other waveforms that might give a higher Rx signal for the same average coil current? Average coil current=average on current*%on time. What would be the best %on time? Any thing to consider before the preamp besides coil diameter and Tx waveform?

Another test comparing 160usec Tx vs 1600usec Tx. Constant rate gives a higher Rx signal than constant current with the same average coil current so I used constant rate only for the test. Four targets, US nickel(TC about 10usec) US quarter(TC about 125usec) 1oz copper coin(TC about 450usec) and California ground(quart zip lock bag, 7x6x1.5inches). Target laying on, centered on one of the figure8 Rx coils. Two record times, 250usec and 1000usec. Charted log-log(easier to see %difference)Some observations: As expected the signal for the nickel was the same with both Tx times. US quarter(signal for 1600usec Tx about twice the signal for 160usec Tx). 1oz copper coin(10/7 higher at 10usec, more than twice at 100usec, decays faster with 160usec Tx time). California ground(decays faster with the 160usec Tx). 1600usecTx(signal is higher for higher TC targets, for the same average coil current Tx rate needs to 1/10 less). Don't know if S/N would be any better(averaging fewer samples and integrator gain resistor ratio or sample time would need to be increased. Recordings not as clean as I would like but if a line is drawn through the noise it should be close to a better recording. Not sure what to expect with the 1 and 1.5inch copper cubes(expected TC maybe between 4000 and 10000usec)

[green]

There is a lot of good information on the internet dealing with 'transient electromagnetic ore body detection' This is going to the other extreme, but the technology is basically the same, but much scaled down in our case. In fact, geophysics is also moving into the 'near surface' applications.

Have a look at the following - Google 'Broadband Transient Electromagnetic Techniques'. Scroll down to the .pdf by Zonge where it talks about 'Improved Methods for Detection and Classification of UXO's'. The merits of the half sine pulse method are discussed and how this results in a greater transmitted moment for the same current draw. You will need to magnify the on-screen text and pictures to see the considerable detail on the poster.

Another paper that is interesting and deals with an unusual coil arrangement is - 'Opposing -coils transient electromagnetic method focused near-surface resolution'. Again, this is a downloadable .pdf.

Eric.

Thanks for the links. Tried generating the half sine in spice. Tx current and Rx signal look like I think they should. Including the zip files and a spice picture. With R2=1(TC=1usec) both examples, same Rx signal. With R2 .01(TC=100usec) both examples, constant rate better than half sine. Maybe I'm thinking or doing something wrong?

[Ferric Toes]

I have been conducting some practical tests on finding the optimum TX pulse widths for particular objects. I set up a board with a TX which I can vary between 500uS and 4000uS. The TX current remains constant at 3.0A flat topped between these limits by dropping the pulse frequency proportionally. The mono coil is 5.5in diameter L = 325uH and R = 4.8ohms with about 10in of coax to the pcb. RX is a bog standard NE5534 set for a gain of 470. Current draw from a 15V psu is 1.3A. This is high, but fine for bench testing, although the coil gets hot enough to melt hot glue adhesive if left on for more that 10 minutes. Good job I used high temperature wire! The decay signal and TX width were measured on a scope with variable cursors.

All the objects were placed in the coil such that the RX saturated for the first part of the waveform. The waveform was then examined at its far end where it disappeared into the 0V base line. This gave the length of the total decay up to 99% or 5Tau. Only the tiniest movement was observable at this point where the object was in the coil, or not.

Test objects were also measured on a Hocking conductivity meter for IACS%

All the objects tested will be tabulated another time but three were selected as being relevant for deep detection of higher conductivity objects. No detection ranges were attempted at this stage as there is no working integrator/dc amplifier on this test pcb. The sampling timing and integrator TC will without doubt need to be optimised in a further set of experiments. Additionally, at least a 15in coil would be used in practice.

1. 1964 US silver quarter$ - IACS 87% - Total decay time 1500uS - Optimum TX width 550uS - Pulse rate 736.

2. 1964 US silver half$ - IACS 90% - Total decay time 2000us - Optimum TX width 750uS - Pulse rate 537.

3. 1oz 0.999 fine silver medallion - IACS 102% - Total decay time 4000uS - Optimum TX width 1360uS - Pulse rate 296.

The optimum TX width was determined by examining the time that the RX came out of saturation with the target in place. Above the optimum time, any increase in TX width had no effect. Below the optimum time the saturation time got shorter, indicating that the received signal was decreasing in amplitude.

Pulse rates could be halved to make the current draw more acceptable.

One interesting finding is that a later 1993 $quarter with the copper core has an IACS% of 41 but the total decay was 1400uS which is not far different from the silver one.

Eric.

[Skippy]

On a (13kHz) VLF machine, clad 25c and 90% silver 25c coins read almost identically, with a time-constant TC = 100 usec (f_c = 1600Hz). The Hocking machine runs at a freq of 250kHz, I recall, so predominantly reads the skin, less so the bulk core.

Will Eric's next experiment require a tungsten wire coil encased in ceramic to stop it melting? It could be handy underwater - the film of boiling water vapour on the coil surface would reduce drag when sweeping....

[green]

I have been conducting some practical tests on finding the optimum TX pulse widths for particular objects. I set up a board with a TX which I can vary between 500uS and 4000uS. The TX current remains constant at 3.0A flat topped between these limits by dropping the pulse frequency proportionally. The mono coil is 5.5in diameter L = 325uH and R = 4.8ohms with about 10in of coax to the pcb. RX is a bog standard NE5534 set for a gain of 470. Current draw from a 15V psu is 1.3A. This is high, but fine for bench testing, although the coil gets hot enough to melt hot glue adhesive if left on for more that 10 minutes. Good job I used high temperature wire! The decay signal and TX width were measured on a scope with variable cursors.

All the objects were placed in the coil such that the RX saturated for the first part of the waveform. The waveform was then examined at its far end where it disappeared into the 0V base line. This gave the length of the total decay up to 99% or 5Tau. Only the tiniest movement was observable at this point where the object was in the coil, or not.

Test objects were also measured on a Hocking conductivity meter for IACS%

All the objects tested will be tabulated another time but three were selected as being relevant for deep detection of higher conductivity objects. No detection ranges were attempted at this stage as there is no working integrator/dc amplifier on this test pcb. The sampling timing and integrator TC will without doubt need to be optimised in a further set of experiments. Additionally, at least a 15in coil would be used in practice.

1. 1964 US silver quarter$ - IACS 87% - Total decay time 1500uS - Optimum TX width 550uS - Pulse rate 736.

2. 1964 US silver half$ - IACS 90% - Total decay time 2000us - Optimum TX width 750uS - Pulse rate 537.

3. 1oz 0.999 fine silver medallion - IACS 102% - Total decay time 4000uS - Optimum TX width 1360uS - Pulse rate 296.

The optimum TX width was determined by examining the time that the RX came out of saturation with the target in place. Above the optimum time, any increase in TX width had no effect. Below the optimum time the saturation time got shorter, indicating that the received signal was decreasing in amplitude.

Pulse rates could be halved to make the current draw more acceptable.

One interesting finding is that a later 1993 $quarter with the copper core has an IACS% of 41 but the total decay was 1400uS which is not far different from the silver one.

Eric.

Interesting test. Wondering if you could add a couple data points the next time you test.

1964 US silver quarter$ - IACS 87% - Total decay time 1500uS - Optimum TX width 550uS - Pulse rate 736.

Saturation time at optimum Tx pulse width _______ and Tx pulse width when Rx amplitude is 1/2 saturation voltage at the recorded saturation time ________.

Example:

1964 US silver quarter$ - IACS 87% - Total decay time 1500uS - Optimum TX width 550uS - Pulse rate 736- Saturation time 960usec - Tx width(Rx 1/2 saturation volts @ recorded saturation time) 185usec