Shielding the sides of a coil.

[ATX_metalprototype]

Hi all,
I am trying to build a detector to detect metal objects above, below and passing through the coil - kind of the idea behind this forum, but I am trying to ignore objects that pass by on the outside of the coil. Is there any way to shield/limit the effectiveness of the field on the outside of the coil? I am using a SurfPI kit and my coil is ~12" diameter and 17 turns.

Thanks!

[Carl-NC]

It's close to impossible to "shield" against detection. Instead, you design the coils to minimize outside detection. For example, a coaxial stack with RX+|TX|RX- will detect objects as they pass through the center or across the face, but outside metal that is fairly static will present a balanced signal. Walk-through security detectors use a figure-8 coil for the RX which works in a similar manner. These techniques do a fair job of ignoring outside metal as long as the outside metal isn't sweeping across the balanced coils.

[ATX_metalprototype]

Interesting, that makes sense. I was wondering how walk-through detectors are configured... Is RX+|TX|RX- a PI configuration, or is it considered induction balance?

Thanks!!

Edit: Picked up your book and answered my own question. It sounds like using the RX+/- method is a pretty drop in replacement to any PI system. And in case of the timing, it looks like on the Surf PI the timing isn't pulled right off the coil rail, so it should be pretty easy to splice in a RX+/- coil and keep the rest of the circuit intact. Or am I misunderstanding the complexity?

[green]

Interesting, that makes sense. I was wondering how walk-through detectors are configured... Is RX+|TX|RX- a PI configuration, or is it considered induction balance?

Thanks!!

Edit: Picked up your book and answered my own question. It sounds like using the RX+/- method is a pretty drop in replacement to any PI system. And in case of the timing, it looks like on the Surf PI the timing isn't pulled right off the coil rail, so it should be pretty easy to splice in a RX+/- coil and keep the rest of the circuit intact. Or am I misunderstanding the complexity?

Some tests I did awhile back. Target was positioned 1 inch above and outside the coil where moving it inwards made no change. Target was moved inward until a 500uv change in integrator out was seen. Recorded distance offset. Repeated for other distances above coil. (coil test, figure eight_2)pendulum number is the distance the target was detected swinging from a pendulum. Detected outside the coil, a lot outside when target was 1 or 2 inches above coil. Rx coils are opposite so a absolute value circuit is needed to detect both Rx coils. Rx opposite so there is a null in the center. Picture of coil before foaming and shielding.

[Carl-NC]

Interesting, that makes sense. I was wondering how walk-through detectors are configured... Is RX+|TX|RX- a PI configuration, or is it considered induction balance?

Thanks!!

Edit: Picked up your book and answered my own question. It sounds like using the RX+/- method is a pretty drop in replacement to any PI system. And in case of the timing, it looks like on the Surf PI the timing isn't pulled right off the coil rail, so it should be pretty easy to splice in a RX+/- coil and keep the rest of the circuit intact. Or am I misunderstanding the complexity?

The coaxial and figure-8 styles are IB and can be used with either VLF or PI. On the SMPI, you would continue driving the TX coil and then add an RX+/RX- coil pair driving the preamp. To do this, you will need to remove the TX coil drive to the preamp. The RX coils will need to be individually damped. It's all conceptually simple but tricky.

[Math]

i have some of that flexible strip magnet ive been meaning to try that round the circumference of a coil to see what effect it has
..has anyone tried this?
another question on coil shielding came to mind recently after working on a fisher 1265x then a tesoro silver sabre..
is there any difference having the shield at 4volts (Fisher) to 0volts (Tesoro) with detector coils operating near ground potential?

[Funfinder]

Reminds me of my "directional MD antenna" thread:
https://www.geotech1.com/forums/show...nal-MD-Antenna

Yes, stacking the coils - Yagi antenna style but with magnetical loops.
Pulsing the signals from coil 1 to coil 5 and amplifying it directionally.

But shielding the coil on the outside is a waste until you don't
reach a significant depth-improvement with it, because that way you
also will cover much less surface-area while sweeping.

However for ferrite-coil-pinpointers its pretty stupid detecting on the
sides instead of just from the tip on forward.

[dbanner]

i have some of that flexible strip magnet ive been meaning to try that round the circumference of a coil to see what effect it has
..has anyone tried this?
another question on coil shielding came to mind recently after working on a fisher 1265x then a tesoro silver sabre..
is there any difference having the shield at 4volts (Fisher) to 0volts (Tesoro) with detector coils operating near ground potential?

The fisher 1265X coil shield is connected to Ground and not to +4v. At least that is what the one of the schematic shows, do you see otherwise with actual coil????? On the other schematic, I see the shield connected to +4V(which is virtual ground or common for this circuit), which is the right way????
Now this is very important observation, as the shield is normally connected to "virtual ground" also called "COMMON". For the 1265 circuit, this is +4v.

[Carl-NC]

i have some of that flexible strip magnet ive been meaning to try that round the circumference of a coil to see what effect it has
..has anyone tried this?

Probably won't do anything.

But shielding the coil on the outside is a waste until you don't
reach a significant depth-improvement with it, because that way you
also will cover much less surface-area while sweeping.

There are other reasons for "shielding." With a security walk-through, we want to detect inside the gate but not outside. At all. Getting zero detection outside the gate is a significant benefit even if it doesn't add any sensitivity inside the gate.

Another solution (mentioned by Davor in the other thread) is a Halbach array. These work similar to refrigerator strip magnets where you get a field on one side but not the other. However, they are a bit complicated to build.

[Math]

hi Dbanner yes thats correct the shield on the 1265x is connected to +4v common.. ive double checked, mine is an early model it had a plastic coil connector and socket fitted but not sure if that had been added at some point later..
ive experimented with the shield connected to ground and theres a sensitivity loss and increased current consumption

hi Carl thanks i have a look at the Halbach array

[Ferric Toes]

One method that I have used to make a PI search head that only detects on one side, is to use a ferrite pot core half. The largest size available seems to be 150mm x 30mm and with a coil wound around the centre pole, one sided detection can be achieved. In one industrial application the back of the core was up against an aluminium plate which only gave a minimal signal that could easily be compensated. If a aluminium dish was made to cover the sides as well as the back then true one face detection is possible. To get more detection range and to be useful for ground searching or doorway detecting, we need Ferrite Lego so that a shielding box of any size or shape can be made.

Eric.

[Funfinder]

These are valueable experimental technical experiences wich could be useful sometimes.

Thing with ferrite is, as wie can see with AM-receiver, that it "pushes" the EM-field
to the outside through its undirectional "anti-magnetic-metal"-structure while an
iron-core at a generator-coil would concentrate it.

Seen from the weight and frequency its pretty useless and
I doubt the ferrite pot will "push" the + field
into the - direction so the detection depth doubles.
Placed on the outside ring of the coil might be a different story.

And making a MD less sensitive or block it at one side by adding
ferrite and metal-sheets doesn't means that the sensitivity
at the contrary side of the coil increases automatically.
It might just decreases anything and adds weight.

But what might be useful is shielding the coil-cable near the coil
with some selfmade ferrite-powder tape.

[Ferric Toes]

These are valueable experimental technical experiences wich could be useful sometimes.

Thing with ferrite is, as wie can see with AM-receiver, that it "pushes" the EM-field
to the outside through its undirectional "anti-magnetic-metal"-structure while an
iron-core at a generator-coil would concentrate it.

Seen from the weight and frequency its pretty useless and
I doubt the ferrite pot will "push" the + field
into the - direction so the detection depth doubles.
Placed on the outside ring of the coil might be a different story.

And making a MD less sensitive or block it at one side by adding
ferrite and metal-sheets doesn't means that the sensitivity
at the contrary side of the coil increases automatically.
It might just decreases anything and adds weight.

But what might be useful is shielding the coil-cable near the coil
with some selfmade ferrite-powder tape.

I suspected that you might be the first to answer. Perhaps you can explain a little more what the 'unidirectional anti-magnetic-metal-structure' is. What we 'see' with an AM radio with a ferrite antenna is not relevant anyway as the type of ferrite material used for this is different and will not work for a core in a PI detector sensor. You also might explain in more detail how having 'self-made ferrite powder tape' on the coil cable near the coil might help in having a more directional field. Does it 'push' the field away from the top of the coil?

Eric.

[dbanner]

Oh dear dear, lord Eric........

[green]

Anyone test a DD coil for side detection? Some test data I did awhile back. Didn't test for side detection, would except zapped my test circuit.