coils

[josephrz]

I have built Mark Stuart's and Gary's PI detectors. Like everyone else 'am trying to find the best coil winding method. Is it: bundle, spiral, basket, cylindrical wound, or yet something else? The target in mind is a gold ring at a Texas beach. My two layer 8" o.d. cylindrical mated to Mark's circuit picks up the specified target at 12" in free air.

Second question: PI guru, Eric Foster, likes stranded wire, is there any real benefit to using litz (it ain't cheap).

[Joe]

Re: coils

Josephrz,
I'm about 99% done with Gary's Pulse detector myself. I made my own 45 winding flat coil about 10" dia. I let you know how it works. Have you read much about basket weaving a coil and what do you think? If, It turns out ok I may make a 10x14 coil next. Let me know how yours works out. HH-Joe

[Beachcomber]

Re: coils

How are you going to shield your flat wound coil? I would thing with a flat wound coil that you will have problems with ground capacitance or did you flat wind it vertically instead of horizontally?

HH

Beachcomber

[Beachcomber]

Re: coils

In one of Eric Fosters post on the PI Forum last year he mentioned that PI's react differently than VLF when it comes to coil size vs target size and that for best detecting depth of a gold ring a larger coil of 11" or more is better. I forget what the exact size he said was optimum for a gold ring. I think it was around 18-20" but I am sure if you post this question on the PI Forum and ask for Eric to respond he usually does.

HH

Beachcomber

[Philippe]

Re: coils -->shield

Hello,
I read in this forum that it is important to shield the coil: my questions are:
1) Does the shield give a better response to target (depth) ? and why
2) I'm ok that the signal has a better shape : less noise but Does the coil have also less sensitivity from target signal?.
3) How the capacitance of the ground make trouble in the target signal?

A lot of questions but i need some "ligth" from GURU of PI.
Thanks a lot
PhilippeM

[Reg]

Re: Lets talk coils

Hi Philippe

Coils are a topic of major concern for most people who build their own detectors. With all the winding options out there for building a coil, it is extremely difficult to determine just what would be the ultimate coil. However, my experience from building coils ranging from 1 mh to about 300 uh and from 8" diam to about 11" diam in the mono styles and up to 11" by 16" in the DD configuration and tells me there won't be tremendous difference as long as the key factors are taken into account. Yes, there will be differences in the minimum delay capability or even some in depth capability due to the different sizes involved but one shouldn't be expecting dramatic changes just due to changing something like the type of wire used, or even the shape of the coil.

There will be small but noticeable changes with any of the factors mentioned, with coil size having the greatest effect on coin or large ring type signals.

One item that needs to be mentionoed is sometimes changes are made such as increasing the coil size and suddenly, there seems to be a large increase in signal strength from a particular object. In many cases, the large increase is usually due to the shift of just when the delay is taken. If the delay is right at the edge or partially up the delay curve, the signal is enhanced and will appear to be much stronger.

1. Now, to tackle Philippe's first question, shielding of the coils and is it important. The answer to this question is yes, especially if you want to use a very short delay and have a very stable detector that doesn't give a lot of false signals.

If you make a coil without shielding, you will find the detector will be very ground sensitive and touchy. You will most likely experience strange false responses when the coil touches certain weeds or other items.

One of the strangest responding PI's I ever encountered was a friend of mine's. He mistakenly tied the shield to the signal lead instead of the ground lead. The detector whined, banged, and made some really strange sounds as the coil approached most anything.

This mistake really emphasized just how important the shielding was. One should try it just once to see.

2. Does the shielding reduce the sensitivity? My best guess is yes if tested in a perfect controlled environment where there was no noise and using an air test, but I don't expect the loss to be not that much. However, there is no perfect environment and people seldom try to detect flying objects (which is about all one could detect in an air test), so overall, the shielding would have a real effect of increasing the sensitivity simply because one can better distinguish a target from a ground signal or from noise.

There are three major signal problems encountered with a PI. One is the ground signal, another is false signals from weeds and grass, and the other is the noise. All can effectly reduce the sensitivity or useability of a PI dramatically. If one of the problems doesn't affect the detector , one of the others probably will. Now, the shield can effectively reduce all three problems.

3. Your third question has been basically answered. The ground capacitance will significantly increase the ground response which makes it difficult to keep the PI properly "tuned" for the greatest sensitivity.

Ideally, one wants the detector to be operating such that the slightest change can be heard. If the ground is causing significant signal changes, then it is impossible to distinguish a very weak target from a ground change.

Coil Size: One can build coils about any size they want for a PI. As a general rule, the larger the coil, the bigger the antenna. So, as the coil size goes up, the more noise one will encounter.

On the plus side, the larger the coil, the deeper one can detect some objects. The larger coil produces a field that simply goes deeper, so it is possible to detect larger items deeper with a larger coil.

On the down side, the larger the coil, the less sensitive it will be to very small low conductive objects buried deeper. This would include small gold nuggets or maybe very thin gold chains.

Another down side of a larger coil is the fact that it becomes more difficult to make the coil work at very short delays due to the coil capacitance. One can reduce the inductance which then will allow better detection of the smaller gold items.

The delay is another topic but since coils have a direct bearing on the minimum delay possible, it is probably a good idea to discuss it briefly. Not only does the size of the coil have a bearing, but the number of turns will help determine the minimum delay that can be used.

More windings make a coil go deeper, right? Not necessarily. It depends upon what a person wants to find. If gold rings are the objective, then one wants a detector with a shorter delay. Large gold wedding bands are not a major problem, but thin gold rings, the type that usually have diamonds mounted on them can be. There isn't much gold in such a ring, so a short delay is necessary.

A 1 mh coil may only allow a minimum delay of 15 usec to 20 usec, while a 300 uh may allow delays down to 10 usec or so. (It does take some real doing to get a coil to work at 10 usec.) Now, one should keep in mind that all signals are stronger at 10 usec than at 20 usec, so the shorter delay can enhance the depth more than the increased signal field strength from a coil with greater inductance.

Finally, mono coil versus a DD coil. A mono coil has the real advantage of a full detection zone across the coil. A DD coil will primarily detect a target when the target is under the overlap zone.

A mono coil is more ground sensitive and noisier than a DD coil. The DD design does help reduce the ground response, but also seems to reduce the noise pickup also. Part of the reason for the noise reduction is the windings of a DD coil are smaller. Another part is the fact the receive coil is isolated from the transmit portion of the detector. The result is, a similar size DD coil can, in some instances, go deeper simply because of the two mentioned factors.

Reg

[Reg]

Re: coils

Hi Josephrz,

I built Mark Stuart's PI some time back. I made several mods to the circuit to help stabilize the thing and added more gain and an autotune feature which also helped. My objective was to try to build a PI that would respond to very small gold. As I kept modifying the detector, I finally ran into the earth field effect signal and PS noise. At that point, I decided it wouldn't fit my application.

I also found the detector would respond strongly to certain "hotrocks" so I ended up reducing the delay down to about 80 usec to minimize this problem.

Fortunately, I doubt you will have that type of problem on a Texas beach and the search for gold rings is quite different than trying to find very small gold nuggets, so the detector should work for you ok.

Now, I doubt you will see any advantage at all by switching to Litz wire. I have used Litz wire and still do on some of my coils simply because I got it cheap.

You might check out the prices at: http://www.surplussales.com/

Just select product index, then wire, then Litz.

Hopefully, this next link will take you directly to the page where the Litz wire is displayed.

http://www.surplussales.com/Wire-Cable/LitzWire.html

What they have is surplus and when it is gone, it won't be available at the prices mentioned. You probably would have to use the largest size they have though because of the current capability. Their served wire is about equiv to something between a 25 awg and a 26 awg.

I also doubt you will see any sensitivity difference between any of the different winding configurations you mentioned if nothing else is changed.

The advantage of using different types of wire or even winding configurations is to reduce the capacitance of the coil. This will allow the use of a higher resistance damping resistor, which then will allow the use of a shorter delay.

A shorter delay should produce a stronger sample signal from an object simply because the signal from that object hasn't dissipated as much.

Now, I doubt that changing either the winding configuration or wire itself will result in a tremendous change. I have found I can change to plain stranded or certain Litz and reduce the delay maybe a usec or so. I suspect the winding configuration might help a little, but not much. But, then again, I haven't tried it either. I would be interested in the results if anybody has.

The trick for maximum sensitivity is to set the delay as close as you can to the point when the decay is at 0V and getting the decay to occur as fast as possible. This combination will result in sampling the strongest signal from an object.

Now, since your objective is a gold ring then a long pulse on time is of little value. So, I would recommend you reduce your pulse on time down to less than 100 usec, maybe as low as 50 usec or so. On my present PI, I have gone as low as 30 usec pulse on time and see no signal loss from different size gold objects since I am sampling earlier.

You could experiment with the damping resistor also to find the fastest decay time without any oscillation also.

If you just decrease the pulse on time, it might look like you lost ground and the detector is not quite as sensitive. This could occur of you are presently sampling just a little of decay signal and not when the signal is completely at 0V. It all depends upon the decay of the pulse signal and when you take the sample.

In other words, there is a lot of interaction between the pulse on time, the decay time, and just when you take the sample.

As for depth, you might also increase the size of the coil to maybe 11" or so. This should help depth as much as anything.

Again, the real advantage to using something like Litz or insulated stranded wire is to reduce the winding capacitance so the delay can be shortened. Just changing wire and not doing anything else will most likely result in no change at all.

One last point, you might experiment with the sample time also. A long sample time is something like a long delay before sampling. If it is too long then you will be sampling a very weak signal towards the end which could hurt the final results. Also, the longer the sampling time, the greater the possibility of sampling noise. Don't be afraid of taking the sampling time down as low as 5 usec or so when trying different things.

On some detectors, a very short sampling will work great and may show a sensitivity increase, but it my have a negative effect on other detectors depending upon the design itself.


Reg

[Philippe]

Re: Lets talk coils

Hi Reg,
Many thanks for this long answer.

I am trying to make a PI detector with a friend.
As you said , i'll try to make different coil for testing the effect of the shape, the shield ...
As this test will be valid with measurement , I'll put it on the web for the community.
Thanks a lot
Bests regards
PhilippeM

[Frank Wallis]

Re: Lets talk coils

Hi Reg
Great post

One point you did not cover,was regarding the size of conductors and whether that can have any effect on the targets size and depth.

Regards Frank Wallis

[Eric Foster]

Re: coils

Hi Reg and all,

I have not myself wound a coil out of Litz wire, but have tested the Minelab SD coils which are Litz. I think that the main reason for using Litz is that it reduces the cross sectional eddy currents that can be generated in the wire itself, when the transmitter switches off. In other words, high frequency losses are minimised. This is exactly the reason that it is used in some radio frequency coils. In fact industrial balanced coil metal detectors often used Litz wire for this reason, long before Candy's patent on the subject.

The SD2000 series detectors require a very low resistance TX winding, around 0.4 ohms if I remember right, so they need a larger wire cross section. Ordinary stranded wire would be a bit too lossy as the conductors are in contact, rather than insulated.

I tend to use higher resistance coils (above 2 ohm) and find that finer stranded wire is adequate for the job. The inter strand contact having enough resistance to break up the eddy current path. If you substitute solid wire for stranded wire of the same cross section, you can see the eddy current loss signal clearly on a scope. It has the same effect as reducing the damping resistor value. The result is, that you cannot sample as early.

It would be interesting to wind two identical coils, one Litz and one stranded, with the same conductor cross section, and see if the winding capacitance with the Litz coil was less. I would hazard a guess that there would be little in it.

Eric.

[Reg]

Re: Lets talk coils

Hi Frank,

You are right, I didn't cover the wire size and it could make a difference depending upon the circuit design. My apologies. I have been working with a PI designed by Eric where smaller wire size is used, so I have a tendency to forget how much of a difference it can make.

Wire size becomes very important on a VLF xmit coil. If the wire is too small, the detector will not be that efficient, meaning the detector will draw much more current than necessary.

On a PI, wire size can make a difference also for the reasons Eric mentioned. If the design requires a very high current flow, then a wire that has an increased resistance will have an impact.

I recently mentioned a site that had Litz wire but the size is about 26 awg. This size wire has about 4 times the resistance of something like a 20 awg. This could cause a detector to react differently. One would probably have to parallel strands when winding to accomplish the same thing.

An alternative would be to change the gain characteristics of the design to allow for smaller wire, but that may be impractical at times.

I really need to finish building Carl's PI and try different coils using different size wire. However, I do remember when I added resistance into the coil circuitry on another design I built a long time ago, I did experience a sensitivity loss. At the time I was extremely lacking in knowledge of just what was happening.

BTW, Eric did a great job of explaining the wire size in another post.

Reg

[josephrz]

Re: coils

Esteemed gentlemen: Thanks! One of the benefits of this forum is that there is a huge body of technical knowledge here. Another is that it's members so generously share it. Thanks again for the info, guys.

[Reg]

Re: coils

Hi Eric,

One of the things I did when experimenting with one of my DD coils was to try both a receive winding made from non-served Litz and a different one from stranded wire. The wire size wasn't quite the same but very close. The Litz was equivalent to about a 27 awg and the stranded wire was 28 awg. The stranded wire had a teflon insulation which is generally thinner than most insulation. The only reason for using the Teflon was because I had it available.

Now, I found there was very little difference in how the coils worked when comparing the two different windings. The stranded wire did show a very slight improvement which I attributed to having a slightly less capacitance. I attributed the less capacitance to the insulation separating the individual windings. If I remember correctly, the difference was less than 1/2 usec.

Reg

Reg

[josephrz]

Re: coils

Joe - Gary's board is waiting for a gel-cell to show up. Then it goes into a water-proof Pelican 1040 case. Somebody, I think on this forum, was praising the basket weave config and included a link to a Russian site. Well, I don't do Russian, but I was interested. So, I visited a couple of diy VLF radio sites where they favor both basket and litz. This much I came up with: The name of the game is to produce as high a Q, (a.k.a, low winding capacitance) coil as you can for all the reasons elaborated on this thread. As with using litz, basket weave is one of the ways of getting there. And that's pretty much all I know. No, I am not ready to throw together a peg board to make the thing, at least not yet. But, it's a great idea!

[Frank Wallis]

Re: Lets talk coils

Hi Reg
If you did not have the problem with eddy currents,would signal pick up be better with larger conductor.
Regards Frank Walis