How to judge shield effectiveness?

[bklein]

Our first interest is receiving the target response. This is accomplished in a much shorter timeframe than one 60Hz cycle. The 60Hz signal is corrupting the detector's audio detection means. So can't that be dealt with in the circuit?

Another thing I came across was a site that described making a spark gap with copper tape that ESD would spark. This would be picked up by an AM radio for verification. The shield material under test was placed between the spark source and a loop probe. How does this method compare to the AC mains detection method?

Also, what about the capacitive plate/ground issue. There must be a big difference in effectiveness between a coil wrapped in shield material and a cover of some material wrapped completely around the coil - I would think the cover presents much more capacitive plate area to the ground...

[bklein]

Sorry for multiple posts but I'm reading though a bunch of stuff I printed out earlier - as the day progresses.
I just read this article:
http://emcesd.com/pdf/emc99-w.pdf (read about Figures 6-8 )
The earlier article I referenced talked about the trouble necessary to make the shield balanced. But now read this article and it becomes obvious - it is a fruitless task for us. It also becomes apparent to me at least that the shield should not be wrapped around the coil turns. As you move across ground it will cause an unbalanced shield signal that hurts you rather than helps. It seems to make much more sense now to me that our shields should be complete clamshell covers like Minelab has on theirs. Watcha think?

[Davor]

The spark thing sounds more lethal, but it would do it just fine. You simply need a changing E-field of several hundred volts at plate, and that's it. Many detectors are removing 50/60Hz quite effectively, so the spark thing could be better.

I started searching loop and e-field topics with Google and came across this site:
http://www.w8ji.com/magnetic_receiving_loops.htm
How much of what he suggests applies to our coils?
Note what he says about noise source coupling distance from the loop, how the shield is an acting antenna, how the shield gap must be specifically placed so there is differential signal balance. He doesn't seem to cover shield material choice. I wanted to ask him about that but couldn't find his email.

Nice catch. You have a thorough analysis of the loop antennas there. Metal detector loops are mostly designed as direction finding antennas. Looking at the history of MD development the shield was adopted from there without much asking why.

If you haven't any experience in direction finding (sport) on 80m band, you'll be mostly at loss with many of the concepts here, but most of the concepts apply. Coil shielding in direction finding receiver serves a purpose of providing deep nulls in a direction of a transmitter you wish to triangulate. A dipole or a monopole suck in direction finding because they couple with nearby objects and you don't have such nice and deep nulls as with a loop. They also pick lots of E-field noise. A loop has sharp nulls, provided you don't have any E-field pickup. In case of E field pickup a perfect figure 8 diagram gets distorted in a cardioid manner, hence the perfect symmetrical shielding is a must for good triangulation.

Back to metal detectors.
Our toys operate in a very near field. They are also prone to pick microphone-like signal from the electrically charged irregularities (wet grass) that are coupled to the Rx coil as a common mode signal. Shielding prevents such pickup, but it does not prevent reception of far field EMI. In fact, our loops are very good antennas for far field emissions.

Most of the front-ends in metal detectors, Tx and Rx alike, are single ended. That means that one side of a coil si connected to the ground. Such connection picks equally common and differential mode signals. Shield removes the common mode noises, and you are left with a differential mode signals. Exactly the same is achieved with a differential mode front end that suppresses a common mode noise, with no shield at all.

PI Tx is a single ended signal source, hence it has both common and differential mode components. Such coil's average electric potential is at half the voltage applied to the coil. A shield removes the E-field component so the coil passes EMC regulations, but in a process such coil gains a bit on weight, and becomes a wee bit slower. In case a balanced PI Tx is developed, the average electric potential would be zero, and shield would not make much difference.

In all accounts a shield is taking care of a common mode phenomena, and nothing else.

[Ferric Toes]

Sorry for multiple posts but I'm reading though a bunch of stuff I printed out earlier - as the day progresses.
I just read this article:
http://emcesd.com/pdf/emc99-w.pdf (read about Figures 6-8 )
The earlier article I referenced talked about the trouble necessary to make the shield balanced. But now read this article and it becomes obvious - it is a fruitless task for us. It also becomes apparent to me at least that the shield should not be wrapped around the coil turns. As you move across ground it will cause an unbalanced shield signal that hurts you rather than helps. It seems to make much more sense now to me that our shields should be complete clamshell covers like Minelab has on theirs. Watcha think?

In my experience, wrap around shielding is the best provided a small gap is left, usually where the coil wires come out. The shield material (woven copper fabric, or tinned copper mesh) must have a TC at least 5 x faster than the coil circuit TC on switch off. Even so, there will be very little in the way of eddy currents induced in the wrap around shield as the collapsing magnetic field is concentric with the shield material. At no position is there a component of the field that is perpendicular to the shield as in the case of coated coil shells. Another shield material I have used is fine pitch ribbon cable, again helically wound around the coil, all the conductors are joined at one end and grounded; at the other end they are left open. The end result is a helical comb. To hold it in place initially I wind double sided adhesive tape on the coil winding. All these methods are probably too labour intensive for large production, but I firmly believe that this form of shielding is more effective than coated shells. One other benefit is that there is no chance of microphony, which I have observed occasionally in coated shells where the coil can move slightly relative to the shield if banged on a rock or tree stump.

One thing I don't understand is how you can get "an unbalanced shield signal" with wrap around shielding. The shield is like a one turn open circuit coil with the only connected end going to electronics ground. Where you ground the shield does not matter as there is no current, so you could ground it diametrically opposite the gap if you wished.

By the way I have only used the above methods for PI detectors, both monocoil, concentric, and DD. There may be problems with CW IB detectors that do not show up in time domain.

Eric.

[Stefan]

I have been using a wire shield similar to the fine pitch ribbon cable you described since 1996. However I use enamelled copper wire and adjust the spacing distance. With this type of screen you do not leave a gap so the coils is fully screened. The beauty of this screen that it is very robust and you do not get microphony when the coil is bumped against a rock. It is labour intensive to make but well worth the effort.

Stefan

[bklein]

One thing I don't understand is how you can get "an unbalanced shield signal" with wrap around shielding. The shield is like a one turn open circuit coil with the only connected end going to electronics ground. Where you ground the shield does not matter as there is no current, so you could ground it diametrically opposite the gap if you wished.
Eric.

The Magnetic Loop Antennas Receiving article (pages 3 and 5) suggests there is inductive coupling between the shield and the inner conductor of the loop (our multiple turns). I guess if it is split in half by moving the gap opposite the cable junction the influence of this is cancelled out. But that is fine for a common mode ground/e-field influence - but that is not really what we deal with. We move the coil over a variation in the ground that influences one half of such a shield more than the other.

I'm not trying to support any particular view or act like I know what I'm talking about. I don't.
This all would be fun to test out - if I had all the time in the world...

[Mechanic]

Hi Guys,

My take on the gap in the shielding is this..... It is mostly used when the shield material is highly conductive. If you were to join the shield it would form a loop. This would cause eddy currents to form in the shield that would take too long to dissipate. To test this, take a piece of copper wire, form it to a loop, but leave the ends apart. Check the detection distance. Now join the 2 ends and re-test. Do the same with a thinner piece of wire that is barley detectable without the ends joined, then with them joined.

When you use a higher resistance shield material, such as carbon paint you can coat the whole lot without causing any problems.

Cheers Mick

[Davor]

The Magnetic Loop Antennas Receiving article (pages 3 and 5) suggests there is inductive coupling between the shield and the inner conductor of the loop (our multiple turns).

At radio frequencies it is so.

[satdaveuk]

Hi Guys,

My take on the gap in the shielding is this..... It is mostly used when the shield material is highly conductive. If you were to join the shield it would form a loop. This would cause eddy currents to form in the shield that would take too long to dissipate. To test this, take a piece of copper wire, form it to a loop, but leave the ends apart. Check the detection distance. Now join the 2 ends and re-test. Do the same with a thinner piece of wire that is barley detectable without the ends joined, then with them joined.

When you use a higher resistance shield material, such as carbon paint you can coat the whole lot without causing any problems.

Cheers Mick

Hi Mick
I know where your coming from far as eddy currents but its not as simple as that because if your using carbon spray that giving you a high resistant overall shielding its wont be an even continuity, because as your getting towards the termination point your resistance will drastically decrease like a volume pot or preset.
To give you a example I'm just making another 8" coil for the Barracuda, cut the former and routed a 10mm grove round it, and painted the grove with graphite mix all the way round continuously (no gap) which is giving me 5k reading from end to end. Now I am creating a 2mm gap to open circuit it , the reading is 71k , so what im trying to say is the nearer the shielding gets to the termination point where you connect your earth the less resistance you will have at that particular point which in turn will give you a unbalanced coil from beginning to end.
I have have made this statement but now I am doubting myself so if I am wrong please say so and why?
Actually I am making this coil to try Eric's 10k per inch and see how it works over here, have left enough room in the routed grove for a 1mm rubber sleeve before I wind the wire and finish shielding but bit concerned if its wide enough to keep inter capacitance down, if not will have to just strip the lot out and increase the width..

[grumese]

Hello, just wanted to congratulate and thank you all, for everything that I learned from all of you.
My name is Angelo and I am just a beginner in search of knowledge.
I managed to build my coil, following the instructions posted here.
Thank you all

[Pin Head]

I am not sure if you will get this but I was reading up on this thread and was wondering why on pic 4 you have a thin wire around your coil is that for shield and did not use graphite? Our did you use both? I think this went to the wrong thread I was trying to get info from Tepco on a older thread where he had posted pics of a spiral wound coil. I am new at this forum and not sure how to use it correctly.

[Pin Head]

Ok right thread it is on page 1 post #3.

[6666]

One of the methods for making a shield is to wrap a very thin wire around the coil with a spacing of about ~5 -10 mm

[Qiaozhi]

One of the methods for making a shield is to wrap a very thin wire around the coil with a spacing of about ~5 -10 mm

That gives me an idea for an informative (and simple) experiment, if anyone wants to give it a go in their spare time.

Here's the question to be solved: "What is the maximum spacing for the thin [drain] wire (described in 666's post) that still provides effective shielding?"

I suppose we should also consider ways of testing the shield's effectiveness on the bench. Any ideas?

[Koala]

That gives me an idea for an informative (and simple) experiment, if anyone wants to give it a go in their spare time.

Here's the question to be solved: "What is the maximum spacing for the thin [drain] wire (described in 666's post) that still provides effective shielding?"

I suppose we should also consider ways of testing the shield's effectiveness on the bench. Any ideas?

Think to test we need to add something near the coil that will introduce the chatter that we can all use.

I have had a coil with a broken shield (VLF coil IDX) when testing in the house blowing about two foot over the coil was enough to cause the detector to chatter. But the same coil still worked fine on the beach over the dry powder sand.

Would charging a balloon up by rubbing with a cloth be a fair test ?