Current control

[Tinkerer]

In highly FE mineralized soil the inductance of the TX coil's inductance increases with the proximity of the soil. This affects the decay time of the coil.

What would happen if we control the coil current, pulse by pulse, so that we can counteract the inductance change?

Non ferrous targets do not affect the inductance, only the coil current.

I have an idea of how to implement this on the breadboard, but would like to hear opinions.

[greylourie]

Could we not use Moodz automagic damping to control flyback, and hence, control associated current. Excuse me, I am a novice.

Would be nice if an open source solution could be found.

[greylourie]

Hello Tinkerer

The idea you have, would it only be of use for an I.B system ? Or could it be of use with other popular P.I projects found here on Geotech ?

Regards

grey.

[Tinkerer]

Hello Tinkerer

The idea you have, would it only be of use for an I.B system ? Or could it be of use with other popular P.I projects found here on Geotech ?

Regards

grey.

This idea is still on the idea level. It could be interesting to be tried and investigated.
With constant TX time, the amount of current in the coil, increases with the present of a magnetic permeable core or target. This in turn increases the time for the coil current to decay, therefore we see a target.
If we control the damping resistor, pulse by pulse, we can counteract this
With a current sensor, we could set the TX coil current to be switched at a specific current level, overriding the TX timer. The TX time would become variable.

.

[greylourie]

How can we get to produce a constant current ? As the coil moves over varying ground, the inductance is altered... and coil supply parameters are affected too ? So maybe use feedback to adjust ?

This is exciting. I recently found an old thread by B^C, where he posted some graphs taken from a report on the comparative performance of various de-mining detectors. He noted that only one manufacturers product showed close and consistent similarity between in-ground and in-air results. So I think that if B^C s comments had been followed through on, the question you have posed here would have been asked sooner..... And in the Australian forum, Aziz and Mechanic noted this varying of inductance when looking at data from their "Hot3" sample measurements.... and I think we should ask them for some input. Maybe they would be willing to contribute, and throw some ideas around.

Would a solution also provide relief to coils passing through varying resistive surf, beach-side ? Or only to magnetic targets seen by the tx ? If it could, then maybe we could get more members interested in finding an open source solution. Or am I being clueless here.

Maybe addressing the problem right at source, (the coil end) is the final and most crucial key to making a good metal detector. Thanks for posing this question in your thread, Tinkerer.

[Tepco]

I did this before, in some soil sample test setup, even posted schematic somewhere on the forum, will re post if I manage to find it. Simple one chip (UC3843) solution constant peak current control to replace 555 circuit. With coil inductance increase, circuit will adjust (increase) pulse width, to keep peak current at preset value. Difference is, keeping constant current with inductance increase, energy stored in coil (1\2 IxIxL) will increase, keeping constant pulse width it will decrease. This is clearly visible with small coil and large sample inserted in it, but inductance change of normal size coil affected by ground is very small, not tried but I don't expect any significant change or difference. Non-Fe conductive metallic objects will reduce coil inductance not just current.


Actually I used this in one old modified detector, and find this method nice to increase stability of static mode (non motion) machine, eliminating for example, drift due to battery voltage variation during discharge (coil energy can change 30-40% during discharge cycle with constant pulse) but never tried on critical soil, not sure if there will be any difference. I'm not aware of this used in any commercial detector, someone would figure out this long time ago if it is any good. Time ago I find something similar in one uC controlled (I think Bulgarian) detector, but circuit seems to be used only to identify coil type at power up, then detector switches to constant pulse mode.

[greylourie]

Hi Tepco

So are you saying that there is no point in continuing with this line of exploration ? Is it the flyback that needs to be adjusted ? Only one manufacturer seems to have addressed this issue of what the coil sees ( and passes on down the signal chain ), when it encounters awkward ground. Quite clearly the graphs in B^C's old posts show other manufacturers have ignored/ or failed to address this crucial point.

Moodz patent has countered the affect with automagic dampening, his method totally removes this element from the equation. Tinkerers suggestion cant be so rapidly dismissed. The solution has to be at the front of the system, not way down the signal chain. Could you post a link to your schematic ? Maybe others can come contribute. I sadly lack on the electronics side.

[Tepco]

Circuit is here, with description, just used for different purpose. We don’t need to dismiss anything, someone can try this to see if there is any advantage, then we will know.


http://www.geotech1.com/forums/showt...612#post170612

[Tinkerer]

How can we get to produce a constant current ? As the coil moves over varying ground, the inductance is altered... and coil supply parameters are affected too ? So maybe use feedback to adjust ?

This is exciting. I recently found an old thread by B^C, where he posted some graphs taken from a report on the comparative performance of various de-mining detectors. He noted that only one manufacturers product showed close and consistent similarity between in-ground and in-air results. So I think that if B^C s comments had been followed through on, the question you have posed here would have been asked sooner..... And in the Australian forum, Aziz and Mechanic noted this varying of inductance when looking at data from their "Hot3" sample measurements.... and I think we should ask them for some input. Maybe they would be willing to contribute, and throw some ideas around.

Would a solution also provide relief to coils passing through varying resistive surf, beach-side ? Or only to magnetic targets seen by the tx ? If it could, then maybe we could get more members interested in finding an open source solution. Or am I being clueless here.

Maybe addressing the problem right at source, (the coil end) is the final and most crucial key to making a good metal detector. Thanks for posing this question in your thread, Tinkerer.

At first look, it seems simple, just control the TX peak current to stay on a determined level, using feedback and TX timing override.
When taking a close look, it gets quite complicated, like most things do.
With a permeable target or ground, the inductance increases. Increased inductance means slower coil charging= increased time needed to reach the same coil peak current. But this means there are more Joules stored in the coil field, which means longer decay.

So, to really control the TX peak current, means we would have to use a mathematical formula in the feedback, which regulates the coil current so as to maintain a predetermined amount of Joules.

It would seem that the result should be reduced sensitivity to permeable targets.

Anyway, I think it is worth wile to take a closer look.

[Tinkerer]

I did this before, in some soil sample test setup, even posted schematic somewhere on the forum, will re post if I manage to find it. Simple one chip (UC3843) solution constant peak current control to replace 555 circuit. With coil inductance increase, circuit will adjust (increase) pulse width, to keep peak current at preset value. Difference is, keeping constant current with inductance increase, energy stored in coil (1\2 IxIxL) will increase, keeping constant pulse width it will decrease. This is clearly visible with small coil and large sample inserted in it, but inductance change of normal size coil affected by ground is very small, not tried but I don't expect any significant change or difference. Non-Fe conductive metallic objects will reduce coil inductance not just current.


Actually I used this in one old modified detector, and find this method nice to increase stability of static mode (non motion) machine, eliminating for example, drift due to battery voltage variation during discharge (coil energy can change 30-40% during discharge cycle with constant pulse) but never tried on critical soil, not sure if there will be any difference. I'm not aware of this used in any commercial detector, someone would figure out this long time ago if it is any good. Time ago I find something similar in one uC controlled (I think Bulgarian) detector, but circuit seems to be used only to identify coil type at power up, then detector switches to constant pulse mode.

Unregulated power supply for the TX will have to deal with very large differences in TX currents between a fully charged and close to dropout LI-ion battery. Even with old style batteries the voltage drop is significant, but the detector manufacturers, (like all manufacturers) produce a product that is good enough for the market, never an ideal product. Basic Market rule.
I have used PI detectors 35 years ago and have observed a large improvement in the technology in this time. However, I believe there are still many avenues to explore and many improvements possible, even if it is just for the sake of the art.

For the very aggressive soil in many gold bearing areas, it could make a significant difference.
<Non-Fe conductive metallic objects will reduce coil inductance not just current > I agree for large targets, but probably there is a proportion of coil area to target area where this becomes negligible.
An interesting special situation is the underwater detector in very conductive seawater.

[Tinkerer]

Circuit is here, with description, just used for different purpose. We don’t need to dismiss anything, someone can try this to see if there is any advantage, then we will know.


http://www.geotech1.com/forums/showt...612#post170612

Thanks for the link, very interesting

[Tepco]

Main question here is, what amount of inductance change can be expected in worst possible ground condition, 0.1%, or 1%, and will this worth additional complications in design, what net improvement can be?

[Tinkerer]

Main question here is, what amount of inductance change can be expected in worst possible ground condition, 0.1%, or 1%, and will this worth additional complications in design, what net improvement can be?

I find that IB coils give a great advantage, but one problem they have, is that very aggressive soil offsets the balance too much. I have tried several ways to correct the IB coil balance in the field and am still looking at better ways. TC pulse current control could maybe work.

[geoscash1]

Wouldn't this fall into the same line of thinking with the project PI5 in the book ITMD?

Instead of using the industry standard "sampling and integrating the decay" design, why not use the approach that is mentioned in the
book, use the decay to control a pulse width modulation circuit so the "presence of metal changes or alters the decay
of the coil flyback". Wouldn't that get you into the ballpark with a base idea to get what your after?

I have often wondered (this is an abstract way of thinking I know) why everyone is so stuck on "detecting the object" first, why hasn't the idea
been to first "Un-detect" the soil condition first then find the object within it? It seems to me, the easy part is detecting the metal or object were all looking for
but why not remove all ground effect first then the objects hidden with in it could be distinguished fairly easily in the after circuitry, so to speak.???

I have always been the type of "the glass is half empty", so fill it type of thinker, why not reverse the thought here and get rid of the real problem first then
work on distinguishing what that problem hides, the objects we all search for with in it.. I don't know, maybe this has already been hashed, tried, pursued and
just left as is and is exactly why everyone's focus is on the second portion, the metal objects within the problem.

Some of this stuff can sure make your head spin thinking on it.....

[Qiaozhi]

I have always been the type of "the glass is half empty", so fill it type of thinker, ...

An optimist would say the glass is half full.
A pessimist would say the glass is half empty.
But ... an engineer would say the glass is twice as big as it needs to be.