gold nugget simulation?

[green]

I find most PIs will hit 2-3 grain gold nuggets @ 1-2". Let's say 3-4 grains @ 3".

#8½ shot is ~ 1 grain
#6 shot is 2 gr
#4 shot is 3.3 gr
#3 shot is 4.4 gr

So #4 shot on a really good PI, #3 shot in a bad day. #3 is hard to find. 8½ also. This is why I prefer solder blobs.

Thanks for the reply. I've been forgetting one of the most important things. Detection distance for the target with a good detector(2 to 3inches for #4shot with a good PI). The target(piece of wire, solder blob, whatever doesn't mean much without a distance reference). Need to get some #4 shot. Is there distance reference for other shot sizes or targets that would be good for nugget simulation?

Charted more data
First chart is with targets laying flat. 10 and 18grain nuggets similar decay slope as 19mm of AWG16 copper wire, nuggets had higher amplitude. 4grain nugget, similar slope and amplitude as 19mm of AWG19 copper wire.
Second chart is with targets on edge. 10grain nugget similar to AWG16 wire. 18grain nugget similar slope with higher amplitude than AWG19 wire. 4grain nugget higher amplitude with slower slope than 19mm of AWG22 wire.
Third chart compares 1x1inch aluminum foil with fishing split shot labeled #1, .3grams, not the same as Carl's bird shot. Shot(not much different than 4grain nugget on edge)is higher amplitude with a slower slope than AWG22 wire. Maybe #4 bird shot would be closer to AWG22 wire? 1x1inch aluminum foil is over 300 times the amplitude with a similar slope as the fishing shot.

I'll try to get some lead shot to try. Any suggestions on other sizes besides #4bird shot?

Interested in what Skippy comes up with to compare with other targets.

[Ferric Toes]

A great piece of work Skippy and well done.

I have a gold U.S Double Eagle $20 coin and my Hocking conductivity meter gives a reading of 15% IACS. I also tried a George 5th Half Sovereign and this reads 16.5%, which is what I would expect. Unfortunately, I do not have a nugget with a flat, or large enough, surface for the Hocking to give a reading. However, I selected a flattish nugget close to 1gm and weighed it on more accurate scales where it reads 0.87gm. I then plotted its decay with my PI viscosity meter from 10uS delay at 5uS intervals to 50uS, where the signal virtual disappeared.
Next, I plotted one of the 4mm phosphor bronze ball bearings that I have. It matched almost exactly the decay of the selected nugget. Next, I raked through my washer box and measured several small plain brass washers until I found the closest match. This was a M4 washer and again I have plotted the result. The TC, which is the 33% of maximum amplitude, comes out at around 16-17uS for all three samples.
Detection range for the three samples was measured on a Vallon connected to a 7" coil. Nugget gave 6"; ball gave 4.5", and the washer 6.5".

Eric.

[green]

A great piece of work Skippy and well done.

I have a gold U.S Double Eagle $20 coin and my Hocking conductivity meter gives a reading of 15% IACS. I also tried a George 5th Half Sovereign and this reads 16.5%, which is what I would expect. Unfortunately, I do not have a nugget with a flat, or large enough, surface for the Hocking to give a reading. However, I selected a flattish nugget close to 1gm and weighed it on more accurate scales where it reads 0.87gm. I then plotted its decay with my PI viscosity meter from 10uS delay at 5uS intervals to 50uS, where the signal virtual disappeared.
Next, I plotted one of the 4mm phosphor bronze ball bearings that I have. It matched almost exactly the decay of the selected nugget. Next, I raked through my washer box and measured several small plain brass washers until I found the closest match. This was a M4 washer and again I have plotted the result. The TC, which is the 33% of maximum amplitude, comes out at around 16-17uS for all three samples.
Detection range for the three samples was measured on a Vallon connected to a 7" coil. Nugget gave 6"; ball gave 4.5", and the washer 6.5".

Eric.

[ATTACH]44725[/ATTACH]

Thanks, more numbers to compare.
I've been calculating TC by dividing time to decay 1decade by 2.3 which calculates closer to 5usec for your targets(12/2.3 your graphs).

[Skippy]

Regarding the lead-free solder alloys. There's endless different types, and their electrical behaviour does differ. But a very common, simple alloy is 99.3% Sn / 0.7% Cu, generally referred to type '99C' or '#244'. This has a resistivity of 126 x 10^-9 Ohm.metre, which works out as 13.7% IACS conductivity. I think this alloy would be a suitable trial alloy.

Re: the pure lead, I guess modern clean lead would be most likely to be pure. I intend using good-looking stuff, the unknown Medieval or Roman lumps I've dug up could have all sorts of tin/silver etc in them, ruining their electrical characteristics.
I was thinking it would be easy to make round-ish blobs of these, with the correct weight + a bit, and then squash them in a bench vice until the correct thickness is achieved. Then trim the edge with a file/abrasive paper until the correct weight or diameter is attained.

Re: the Cupro-nickel CuNi25. I calculate your US 5 cent is 1.6mm thick, which is close enough for this test to the Nugget7 model 1.7mm. But as you don't have any other common CuNi coins, you're going to have to source your thicker metal from abroad. Our old, large 50 pence coin ( pre-1994) is about 2.1mm thick, and I can confirm the Aussie 50 cent is CuNi25, and they are more like 2.5mm thick. So either of these big lumps will be OK for the 2mm thick CuNi models.

If you're going to use weight as a way of getting the size correct, the weights of the dummies in lead and 99C solder are:
Lead:
Nugget 7: M = 0.18 gram; Nugget 8: M = 0.44 gram; Nugget 9A: M = 0.87 gram

99C solder:
Nugget 7: M = 0.12 gram; Nugget 8: M = 0.29 gram; Nugget 9A: M = 0.56 gram

I'm assuming that making these dummies will take some time.
I don't know if it's worth 'designing' smaller dummies yet, presumably some folks are wanting grain-sized targets, so want test their machine out repeatably. It's better we start off slowly.

For interest value, here is the 'Fake Dollar' thread on Tom Dankowski's forum.
Sadly, the link to the electrical properties of alloys no longer works, and I'm not sure if I saved it anywhere, bloomin internet.
http://www.dankowskidetectors.com/di...ad.php?2,27267

[Skippy]

Yay to the Wayback machine. Here's an archived version of the 'solder properties' page:
https://web.archive.org/web/20140528.../solder_alloys

[green]

In the absense of a reply, I've gone ahead and made some estimates, so we can get going on this.

Nugget 7 was modelled as a disc 3.5 mm diameter, 1.7 mm thick.
Nugget 8 was modelled as a disc 5.0 mm diameter, 2.0 mm thick.
Nugget 9 was modelled in two ways:
9A: a disc 7.0 mm diameter, 2.0 mm thick.
9B: a disc 4.5 mm diameter, 2.0 mm thick.

These have volumes of 0.016, 0.039, and 0.077 cm³ for nuggets 7, 8, 9A.

9A assumes the area, of 12 x 4.5 mm is circular.
9B works on the idea that the 4.5 mm dimension is the one limiting eddy-current flow, and the length of 12 mm mainly contributes to the strength of response. This is much like Green's copper wire tests - the small dimensions of the wire diameter dominate.

Now for the Time-Constant modelling.

In previous threads on modelling, we tested square targets, such as aluminium drinks-can sheet, and we came up with a formula showing how Time-Constant, dimensions, and electrical conductivity were related:

T-C = 0.072 x D x T x %IACS

Where T-C was in microsecs
D was edge length in mm
T was thickness in mm
%IACS is the conductivity relative to a figure of 100% for pure copper.

This needs to be modified to suit circular targets.
I did some simple mathematical modelling of L and R for a square vs circle, and came up with an approximation:

A square target 1.00 units across is equiv. to a circular one 1.205 units diameter

and so:

A circular target 1.00 unit diameter is equiv. to a square one 0.83 across.

This results in a first attempt circular target model:

T-C = (0.072 x 0.82) x D x T x %IACS = 0.060 x D x T x %IACS

I then tried this out on real targets. I chose cupro-nickel coins, partly as a hunch, partly as I have more confidence in my T-C values, the lack of skin-effect means my 13kHz VLF is more accurate. Coins tried were US 5c 'nickel', UK large 50p, small 50p,and 20p.
I'll save the details for another time ... but they were pretty close, indicating the '0.060' figure in the model was in the 0.05 to 0.06 range.
So for the purposes of nugget modelling, I'm assuming the 'disc' model is:

T-C = 0.055 x D x T x %IACS

Applying this to the physical nugget models, and using the measured time-constants Green quoted for 'flat-on' and 'edge-on', I calculated %IACS figures of:

Nugget 7: T-C = 2.9 / 1.9 us; %IACS = 8.9 / 5.8 %
Nugget 8: T-C = 6.6 / 3.6 us; %IACS = 12.0 / 6.5 %
Nugget 9A: T-C = 5.4 / 2.9 us; %IACS = 7.0 / 3.8 %
Nugget 9B: T-C = 5.4 / 2.9 us; %IACS = 10.9 / 5.9 %

Points to note:
* These are LOW, down in the lead / cupro-nickel range.
* The long nugget9 is best modelled as a small one, the conductivity figures match the other two's better.
* My hunch was based on work done making 'dummy dollars' over on Dankowski's Forum. These are replica US gold 1 Dollar coins, which are 0.900 fine, but have %IACS figures about 15%. My dummies were made from lead-free solder, close to pure Tin.
* Though Carl M. never backed up his reasons for using lead shot and solder blobs (he should've), I felt he was close, based on my tests with lead-free solder, and the fact he should know what he's doing ....

So ... how to make some trial targets ?
Suitable metals that are easily obtained, and their %IACS figures include:

about 14% : 10,20,50 EuroCent coins, aluminium-bronze
about 14% : some lead-free solders
11.5% : 60Sn / 40Pb electronics solder
8.4% : lead
about 8% : UK 20 pence coin, Cu-Ni, (16% Ni)
5.3% : CuNi25 coins, eg. US 5c, UK 50p large / 50p small

I reckon lead-free solder ; pure lead; and CuNi25, covering the top/middle/bottom of the range, would be a place to start.

I'll think about the details and post tomorrow.

Attached pics of the models:

Same as Eric, great piece of work.

Tried your formula with some test pieces I have. Foil pieces measured close to Reynolds specifications so I used them, searched 1oz pcb thickness. With probable measurement errors, calculated multiplier was closer to being the same with the different targets than I expected. Not suggesting multiplier number is more accurate than yours, just what I got. Our TC numbers are always a little different, wonder if it's because your measuring with Tx on and I'm measuring with Tx off. I didn't see much difference in TC, square vs round, some amplitude difference. Thanks again for the reply. I'm going to try some of your targets, maybe in a day or three. Might try 1oz(2inches square)pcb and 2oz(1 and 2inches square)pcb to check against your formula.

first chart 1inch regular Volt should be 1inch regular square.

[Ferric Toes]

Thought I would test on the Hocking IACS meter a couple of items mentioned above. 1" x 1" x 0.2mm lead foil was too low to read. Double thickness (0.4mm) read 2.28. Regular foil of 0.02mm gave no reading for 1 layer. However, 2 off 1.81. 3 off 3.59. 4 off 5.74. 5 off 8.85. 6 off 11.9 Finally the 6 off folded to give 12 layers read 30.3. My Hocking is the high frequency version running at 250kHz. Other reference objects -
US copper penny 48.8
Nickel 5.34
Quarter 40.8

Eric.

[green]

Thought I would test on the Hocking IACS meter a couple of items mentioned above. 1" x 1" x 0.2mm lead foil was too low to read. Double thickness (0.4mm) read 2.28. Regular foil of 0.02mm gave no reading for 1 layer. However, 2 off 1.81. 3 off 3.59. 4 off 5.74. 5 off 8.85. 6 off 11.9 Finally the 6 off folded to give 12 layers read 30.3. My Hocking is the high frequency version running at 250kHz. Other reference objects -
US copper penny 48.8
Nickel 5.34
Quarter 40.8

Eric.

Does the lead foil read on your PI viscosity meter? Maybe a 1inch square of 1oz pcb.

[Skippy]

I've just measured up a few other coins.
The British 2 shilling from 1947 - 1967, and its decimal equivalent, the large 10 pence 1968 - 1992 are another large CuNi25 coin that's just suitable for the 2mm thick nugget models, I measured a few as 1.95mm thick.
And I examined my 10,20,50 Eurocents. The 50c is about 2.1mm thick, so would be a possible choice, but the 20c measures as 1.95mm, so would be good enough. The 10c is 1.7mm thick so only useable for thinner/smaller nuggets. I haven't actually measured any of these 3 coins for TC etc, I will do so soon, as the 14% IACS figure I found online is a bit vague.

[Ferric Toes]

Does the lead foil read on your PI viscosity meter? Maybe a 1inch square of 1oz pcb.

Yes the lead foil does read, particularly on the 10uS delay setting. The tests I have done so far are with the gain and target distance set to give a reading of 1000 (mV) on the display. The target is then removed and next delay is set. Between each measurement the electronics is set to zero by means of a pushbutton. This removes the effect of any spurious or non-linear responses in the RX circuit. As you can see the decay is very fast, such that by 20uS there is no measurable signal. Plotting on linear axes enables the decaying voltage to be easily read such that the 37% point can be determined with accuracy. This measurement corresponds to 1 tau which is read off the X axis to be almost 11.6uS for the single layer.

For the two layer sample the decay becomes longer as can be seen in the second graph. The value of tau increases to 12.7uS and adding further layers would give corresponding increases. It should be noted that electrical contact between layers is not necessary.

As to detection range, the only detector I have set up in my workshop at present is a Vallon VMH3CS with a 7" coil. The single lead foil is detected at 4" and the double at 9". Unfortunately, I don't have any blank pcb. I have a board with a 1/2" track along one edge and that gives a reading of 5.8 % IACS.

Eric.

[green]

Yes the lead foil does read, particularly on the 10uS delay setting. The tests I have done so far are with the gain and target distance set to give a reading of 1000 (mV) on the display. The target is then removed and next delay is set. Between each measurement the electronics is set to zero by means of a pushbutton. This removes the effect of any spurious or non-linear responses in the RX circuit. As you can see the decay is very fast, such that by 20uS there is no measurable signal. Plotting on linear axes enables the decaying voltage to be easily read such that the 37% point can be determined with accuracy. This measurement corresponds to 1 tau which is read off the X axis to be almost 11.6uS for the single layer.

For the two layer sample the decay becomes longer as can be seen in the second graph. The value of tau increases to 12.7uS and adding further layers would give corresponding increases. It should be noted that electrical contact between layers is not necessary.

As to detection range, the only detector I have set up in my workshop at present is a Vallon VMH3CS with a 7" coil. The single lead foil is detected at 4" and the double at 9". Unfortunately, I don't have any blank pcb. I have a board with a 1/2" track along one edge and that gives a reading of 5.8 % IACS.

Eric.

[ATTACH]44739[/ATTACH] [ATTACH]44740[/ATTACH]

Hi Eric, thanks for the tests. I'm wondering why you get 11.6usec for a 1inch square piece of lead .2mm thick when I get 5.7usec for a piece .4mm thick.

I'm still confused with reply #167. Where you state a TC of 16 to 17usec for the targets when your chart shows a TC around 5usec. Maybe I'm looking at it wrong?

[Skippy]

Those charts in #167 do seem to show the nugget having a TC of 5.6 microsecs.

I used decay from 300 to 10, over a time of (35.5 - 16.5) = 19 us

ln (10/300) = 3.40; TC = 19 x 10^-6 / 3.4 = 5.59 x 10^-6 secs.

[green]

Hi Eric, thanks for the tests. I'm wondering why you get 11.6usec for a 1inch square piece of lead .2mm thick when I get 5.7usec for a piece .4mm thick.

I'm still confused with reply #167. Where you state a TC of 16 to 17usec for the targets when your chart shows a TC around 5usec. Maybe I'm looking at it wrong?

Didn't see it this morning. Think Eric is giving us a test to see if we are paying attention. He's reading time crossing at 37% but 100% is at 10usec so time constant is time crossing 37%-10usec(time at 100%).

[Ferric Toes]

Didn't see it this morning. Think Eric is giving us a test to see if we are paying attention. He's reading time crossing at 37% but 100% is at 10usec so time constant is time crossing 37%-10usec(time at 100%).

This is the basic diagram on which my value of time constant depends and can be confirmed in many text books on electronic principles.



Time constant does not change wherever you look on an exponential growth or decay. It would not matter whether you start at 5, 10, 15uS after the true commencement, the value would always be the same for a given L/R circuit, which is what a non-ferrous metal target is.



This shows current growth but the same applies for eddy current decay. This example from my 1955 college book is for a 2H inductor and a 4 ohm resistor gives TC of 0.5 seconds. Stepping anywhere along the curve, even though the voltage and time is changing will always result in 0.5 seconds.

Eric.

[Monolith]

This is the basic diagram on which my value of time constant depends and can be confirmed in many text books on electronic principles.

[ATTACH]44747[/ATTACH]

Time constant does not change wherever you look on an exponential growth or decay. It would not matter whether you start at 5, 10, 15uS after the true commencement, the value would always be the same for a given L/R circuit, which is what a non-ferrous metal target is.

[ATTACH]44748[/ATTACH]

This shows current growth but the same applies for eddy current decay. This example from my 1955 college book is for a 2H inductor and a 4 ohm resistor gives TC of 0.5 seconds. Stepping anywhere along the curve, even though the voltage and time is changing will always result in 0.5 seconds.

Eric.

Would it be correct to say that what we see on a PI decay curve is a convolution of the decay curve of the coil plus the decay curve of the skin effect eddy currents, plus the decay curve of the deep eddy currents (for thick targets)? Plus ground or seawater decay curve etc.