Sample delay and signal strength

[Tinkerer]

To find a target that has an eddy current decay time of 10us we need to sample before 10 us, this is obvious. Now, how do we define the 10us decay time?
Is the decay exponential?
How much signal strength is lost for every us?
What is the initial signal amplitude of that signal?
What is a realistic number of gain to obtain an acceptable S/N near the end of the decay time?
At what point does the signal get lost in the noise?
Does the ground signal obey the same rules as above? As an example: consider we have a ground signal of 20mV after 30us decay time, what was its initial amplitude?

Tinkerer

[pml]

To find a target that has an eddy current decay time of 10us we need to sample before 10 us, this is obvious. Now, how do we define the 10us decay time?
Is the decay exponential?
How much signal strength is lost for every us?
What is the initial signal amplitude of that signal?
What is a realistic number of gain to obtain an acceptable S/N near the end of the decay time?

Hi,

A nugget or piece of metal basically consist of a resister and inductor. The current decay in non-magnetic metals is just like an RL or RC time constant decay like this:
http://www.interfacebus.com/RC_Time_...ng_Voltage.jpg

Here's an RL time constant calculator:
http://ourworld.compuserve.com/homep..._Bowden/rl.htm

At what point does the signal get lost in the noise?

You can't really predict signal noise. I just finished some noise tests in a particular area of Los Angeles, but I have not confirmed everything yet. My amplifier itself generates about 1 nV/Hz^1/2. I was using a 2-foot diameter receiver coil of 80 turns. Often the noise would go below 1 nV/Hz^1/2 for as long as a minute and beyond (usually not that long though), but then the noise could eventually drift in the hundreds of nV/Hz^1/2. The city is very unpredictable and noisy. I'd like to test the desert and truly believe that if you go out far enough the noise level will mostly stay below 1 nV/Hz^1/2.

Does the ground signal obey the same rules as above? As an example: consider we have a ground signal of 20mV after 30us decay time, what was its initial amplitude?

That depends how fast the nuggets current is decaying, which depends on the size and shape of the nugget and it's purity and metal type. If it's a piece of iron metal then the initial amplitude could have been just barely over 20mV since iron decays at a slow rate. A 4 gram typical nugget will decay at a fast rate. So after 30us the signal could be hundreds times less than from its start.

Paul

[Tinkerer]

Decay rate of target

That depends how fast the nuggets current is decaying, which depends on the size and shape of the nugget and it's purity and metal type. If it's a piece of iron metal then the initial amplitude could have been just barely over 20mV since iron decays at a slow rate. A 4 gram typical nugget will decay at a fast rate. So after 30us the signal could be hundreds times less than from its start.

Paul[/quote]

Hi Paul,
thanks for the reply.
You mention fast rate for a gold nugget and slow rate for an iron target.
I understand that the decay curve is allways the same, but can be faster or slower. How do you quantify "fast" and "slow"
Is there any information available about the decay rates of different metals or different ground mineralization?

Tinkerer

[pml]

Hi,

Hi Paul,
thanks for the reply.
You mention fast rate for a gold nugget and slow rate for an iron target.
I understand that the decay curve is allways the same, but can be faster or slower. How do you quantify "fast" and "slow"
Is there any information available about the decay rates of different metals or different ground mineralization?

I go by an estimate that I made many months ago-- no idea how accurate it is, but I think it's close enough for government work. They say typical gold has the electrical conductivity close to lead. For a typical 1 inch round nugget I use 20 nH's and 320 uOhms. If you enter that into the RL time calculator that I previously posted then you see that nearly 1.59% of the nuggets current decays after 1 us, 14.79% after 10 us, 27.39% after 20 us, and 38.12% after 30 us.

Paul

[pml]

I just finished some noise tests in a particular area of Los Angeles, but I have not confirmed everything yet. My amplifier itself generates about 1 nV/Hz^1/2. I was using a 2-foot diameter receiver coil of 80 turns.

I should add that the test frequency was 18 KHz

Paul

[Tinkerer]

1" diameter lead target

pml, thanks for the info. 1" diameter lead target is big, I am looking for 0.5 gram gold nuggets, so obviously I need to sample at 5 or 6 us. I am nearly there, just not quite yet.

Tinkerer

[Unregistered]

How far have you got with your tinkering . did you achieve your desired results in getting the delay down to 5us ..
I too wish to get the sample down that low .. but is a little hard this end .. no scope . Live in Australia where there is heavy mineralised ground and nuggets of all sizes
cheers
gef

[Unregistered]

conductivity of lead and gold

Hi,
I don't think lead and gold are any where close on conductivity. JWFisher puts lead weight in the nose on their towfish on the pulse metal detector. With a conductivity of 7, the signal has died away before the detector samples. I would say that aluminum is much closer to gold and might be a better test for comparisons against gold.

http://my.hrsb.ns.ca/Functions/Curri...nductivity.htm

[Tinkerer]

First sample delay

How far have you got with your tinkering . did you achieve your desired results in getting the delay down to 5us ..
I too wish to get the sample down that low .. but is a little hard this end .. no scope . Live in Australia where there is heavy mineralised ground and nuggets of all sizes
cheers
gef

yes, still tinkering. Got the coil shielded and potted in the form. The first sample delay at 5 uS. Tried different cables with different results. then went back to the preamp to try out some more ideas that Eric Foster has posted over the years. One post specifically, dating back about 5 years got me tinkering for a week and the results are just beyond all expectations. Sensitivity is up by 1000%. I still have a lot of problems to solve and dont want to post results before making a solid pcb. On the breadboard things are too fickel. sometimes I go to sleep in the evening with everything running perfect and in the morning when I power up, I have impossible noise all over. Yesterday it happened after a thunderstorm. Just fine before the rain, then the lightning struck the powerline, no electricity for a few hours, then when power came back, the noise level had increased dramatically. Maybe it is the humidity in the air. If I can give you some advice? Get yourself a scope if you want to tinker. Then read ALL of Eric Fosters posts. There is so much information there. you will have years of fun trying out every thing. Tinkerer

[Kev]

Not the full solution...

Hi Tinkerer,
At 5 or 6us the ground starts coming through with maddening persistence, even with a DD coil. I just did an experiment last week which highlights the problem. I can get a solid signal in air from a 0.4 gram piece at about 15cms with a 250mm DD and a delay of 4.8us, but when I put the gold bit in a jar containing goldfield soil the range is reduced by up to a half.
I suspect the gold's response is being attenuated by the negative counter response produced in the coil from the soil's mass.

This soil is typical of decomposed schist, and is rich in smectite, vermiculite and kaolin that tends to mark minerialised zones where authergenic eluvial gold locally occurs.

Sorry Tinkerer, but your next problem will be ground balance, and what usually then happens is that you need to extend the delay out past 8 or 9us for it to work without causing too much attenuation of wanted signal responses. So it may pay to get some of your local goldfield turf and see what you end up with, because you may have already met your limit.

The places where I find the short delays are best is in the creek beds and rockbars, but the bigger gold is up in eroding tertiary and pleistocene terraces and benches, along with the nasty soils!

Cheers
Kev.

[Carl-NC]

This soil is typical of decomposed schist, and is rich in smectite, vermiculite and kaolin that tends to mark minerialised zones where authergenic eluvial gold locally occurs.

Thou speaketh Greek.

[Kev]

Lazy bad spelling

I know my spelling is bad...this will make more sense I hope.

"This soil is typical of decomposed schist, and is rich in smectite, vermiculite and kaolin that tend to mark mineralised zones where authegenic eluvial gold locally occurs."

The point is that a real short delay is not the complete answer to capturing the short decay of small gold, if one is developing a machine to detect in gold areas derived from metamorphic country rock. They will likely find the above minerals or similar, that will give a ferrous type response at very short delays.

Cheers
Kev.

[Tinkerer]

still a long way to go

Hi Kev,

thanks for the input. you are right, I still have a long way to go, the frontend alone does not do it. And I am not even done with that one yet. I have very little knowledge about ground balance so I would apreciate all good ideas that I could try out. I do have some bad patches of soil nearby, but I doubt if they are as bad as yours down under.
Tinkerer