Some advice on coil selection/design, creating, and measuring needed!

[Pete the Builder]

G'day everyone,

I'm trying to figure out what sort of coil to begin with for the Frankenproject. After a lot of reading, here and elsewhere, I've found there's a lot of great information, but it's spread out all over the place. I'm hoping that many answers can be condensed in one small thread that might help future newcomers. I apologise if this seems ignorant or rude, that's not my intention!

Primarily (no pun intended) this will be most like a PI detector, in terms of the front end. My goal will be to switch from mode to mode to take advantage of the benefits of each type (induction balance, pulse induction, and beat frequency oscillation), but I'm starting off with a simple PI, similar to Tink's nice design.

I have read the coils subforum, and I'm still a bit perplexed. There are great ideas there, but I need to start from the very, very basics. And in the coil forum, there are great photos/descriptions of coils I don't want to make, but only fairly rudimentary explicit visualisations of the coil(s) I do want to try winding!

I have a lot of experience winding transformers and "gauss rifle" type coils, and I have a reasonable ability to design and calculate parameters for all those types of coils, but not for a huge air coil!

As I see it, there are roughly three possible types... flat concentric (spiral wound), "jumbled" circular (or variations thereof - elliptical, oval, D, cloverleaf, etc), and "transformer" wound (neatly layered, etc) to choose from, for both TX and RX coils. I believe I understand the relationship between coil size, penetration, and resolution, this is more about what benefits the different types have.

My first question is, what are the broad relative advantages and drawbacks of those coil types? Is there an online resource I could use, if explaining all this to someone like me is too tedious? Any pointers would be most welcome!

Oh yeah - can I mix 'n' match, or do the coils need to be identical in form?

In terms of ventual use, I'm looking mainly for largeish (2-10" or 50...250mm) ferrous material close to the surface (down to about 10-12", 250-300mm), but I don't want to miss out on any non-ferrous finds either. Think "meteorite" (surface), with the ability to tell me if nuggets of 50-500mm are within about 6-8 inches (200mm) of the surface, most likely in moderate to low mineralised ground. (And I'd like to understand how highly mineralised ground would age this choice, so I can use different coils if needed).

My next question then is how do I easily check the resulting coil's specs? An online calculator or JS or something would be incredibly useful, but I don't seem to have much luck finding one that will arbitrarily solve for the missing variable - the ones I've found all expect me to know the inductance, or some other variable that I want to find! I have a great DMM, but it won't test below 50mH! So any tips or pointers would be much appreciated. I don't think I'm smart enough to fully understand Aziz' beautiful graphic analysis, so just a few numbers is fine!

Finally... Could someone provide pointers to coil-winding tips or jigs that I may be able to re-use and especially resize?

I'm no woodworker, and I have only a hacksaw and handsaw, but I can build something if I can see it! All the images and docs I've found online don't really explain how to easily wind a transformer-type search coil, for example. I've tried using the nails/cup screws ideas from the coils forum, and that works great for a jumble coil, but not so much for a planar or transformer-type coil. Interestingly, I have two 25mm thick 330mm square glass plates, which I use for record pressing, but it occurred to me that it'd make a great planar coil jig... But I'm not sure where to begin figuring out how to start such a jig!

Oh, and can anyone help identify a good source for magnet wire in Melbourne? I buy small lots (up to 100g or so) from Jaycar, but I've spoken to their wholesaler (Electus Distribution), and they can't help, as they buy it in already prepackaged.

I'm sorry to post so much words, but I'm hoping most potential questions will be answered, so I don't waste everyone's time!

Thanks to anyone and everyone for any ideas, tips, or pointers.

Pete

[ODM]

There's basic information about the differences of various coil setups vs. pattern and rejection in http://www.geotech1.com/cgi-bin/page...&file=info.dat
Also about induction, but I figure you're friends with fields and eddy currents already with that background

The Hammerhead project pdf has a bit of info about PI coils vs. capacitance and shielding, and there's more in the Misc section of the projects http://www.geotech1.com/cgi-bin/page...e=projects.dat - I have wound some co-centric coils with three plywood circular discs pressed together with bolts and waxed with candle, fastening the windings with shellac. I remember making a thread about it in the coils, or work methods' sections.

Short primers, of course - there's a bit more text on these topics as well as various circuit configurations in "Inside the Metal Detector" book, and even more in various patents if you browse one of various online patent databases. The latter option takes a lot more coffee due to legalese text formatting! http://www.geotech1.com/cgi-bin/page...le=patents.dat

[ODM]

About inductance calculators, I've found http://www.pronine.ca/multind.htm to give realistic results if you work close to wire size. If you want to find the resulting inductance of n windings, you'll have to work a bit backward - trying inductance values until you get the right number of turns for your coil size.

Might be fun to solve one of the better inductance approximations for number of turns...

About soil mineralization - sadly I haven't run into any of such here in Finland, at least any I'd have noticed with the experiments. I've yet to build a detector I didn't take apart or shelf instead of a new project later.

Does Farnell or any other supply house have a representative near your location? They seem to serve people in Australia rather well, or at least from what I've heard in another electronics forum. Direct orders from them might also be an option.

[Qiaozhi]

http://www.geotech1.com/forums/showt...oil-Calculator

[Carl-NC]

First, stick with jumble-wound for now. Until you get up to stratospheric designs you're not going to see any advantage of other types. Nail beds are simple, I even use them at work until I can justify a cut form.

Second, if you want to measure inductance and you don't have a decent inductance meter, stick a large known cap across the coil and see where it rings. Yes, now you need to know the capacitance, but cap meters are more prevalent, or you can buy a 1-2% reference cap. You also need an o-scope, but you need one anyways.

If you really want a tri-mode detector (IB+PI+BFO) then I would suggest a balanced DD or OO loop with matched coils of ~300uH each. The TX driver is gonna be a real challenge... I've done IB+PI but never added BFO. I have ideas for that when you get far enough along.

Don't chomp down on too big a piece of steak, you're likely to choke. Small bites, chew thoroughly.

- Carl

[Tinkerer]

Pete,

http://www.miscel.dk/MiscEl/miscel.html

The miscel is a very useful calculator around metal detector design. I calculate my coils with it.
The TINKERER designs are not traditional PI designs. They are different animals, maybe ahead of their time.
The TINKERER coils are for the TINKERER designs only, not ideal for anything else.

Tinkerer

[6666]

can anyone help identify a good source for magnet wire in Melbourne?

There is a transformer business in Bayswater on the corner of Barry street and another street,
I dont go past there very often but next time I will try and stop and get more info.
I have bought wire from them before.
Whats wrong with simply soldering the ends together of two spools of wire from jaycar ?

Nail beds are simple and work ok for forming simple PI coils, wrap with string then tape.
Styro foam base, 3 inch nails at angle.

I use a Digitech QM-1324 from Jaycar on 20mH for coils in 100-500uH range all good.

(2-10" or 50...250mm) ferrous material close to the surface (down to about 10-12",

Thinking "meteorite"

Try a simple surf PI with 11 inch coil, dont get too hung up on trying to determine the depth on some meter,
if you can detect the target you can dig it.

[Pete the Builder]

Wow... thanks, guys! Give yourselves a pay rise!

I really appreciate all the advice and pointers, that's exactly what I was hoping to find!

I especially understand the need to take smaller bites... I get a bit too excited sometimes, and end up so far ahead of my abilities that I get frustrated and burn out. I ain't doing it this time.

So right now, it's "feet by the fire" time (er, "feet by the airconditioner time"? ) with the online book (shoulda done that first off...Sigh. And I'm a fast reader, so double that!! ), and I'll dig into the links as well. That really helps to put it all together for me.

Again, thank you all for being such good sports about me asking these things. I hope other folks entering the minefield (pun intended) of detector design that end up here (and they will!) can benefit from this as well. But if not, you've definitely helped me!

Cheers,
PtB

[Pete the Builder]

Another stupid question : Would it be true to say that a large metal object deep in the ground would appear indistinguishable from a smaller, shallower sample of the same material?

If (ha!) this is correct, it seems to imply that it should be possible to fairly simply and repeatably distinguish between the two alternatives, simply by increasing or (preferably) decreasing the next pulse's energy content. Is this in the right ballpark? Or am I forgetting something important?

Oh, and one other stupid coil question, if I may... Would it make sense when designing the TX coil to try to get as much flux into the ground as practical? I mean, using a very low impedance coil (think SWG 18 or lower) with a high voltage, high current source? Or is this going to work against me, putting too much energy into the soil and surrounding material, effectively swamping the signal returns I'm looking for? Note that I'm talking about a very precisely controlled pulse or frequency, not just sticking a car battery across some fencing-wire coils! This could result in a much smaller TX coil, couldn't it?

advTHANKSance!
-PtB

[sido]

I use a Digitech QM-1324 from Jaycar on 20mH for coils in 100-500uH range all good.

6666, have the same meter from Jaycar and works a treat. Paid 50 quid for it i think. Dont know its accuracy for L/C, but would not be to far off as a indicator of Inductance. I like it and does the job ok for coil building and doubles up as a DVM.

Sid

[dfbowers]

It can be reasoned that you could distinguish between a pull tab and a soda can just by how "broad" the signal sounds, but I have been fooled more than once. I can't think of how many times I dug a foot deep and hear a narrow sounding signal still in the bottom of the hole. After digging more than a foot deep you can pretty much conclude that it's not a coin sized object any more.

[Davor]

Very true.

[Tinkerer]

Another stupid question : Would it be true to say that a large metal object deep in the ground would appear indistinguishable from a smaller, shallower sample of the same material?

If (ha!) this is correct, it seems to imply that it should be possible to fairly simply and repeatably distinguish between the two alternatives, simply by increasing or (preferably) decreasing the next pulse's energy content. Is this in the right ballpark? Or am I forgetting something important?

Oh, and one other stupid coil question, if I may... Would it make sense when designing the TX coil to try to get as much flux into the ground as practical? I mean, using a very low impedance coil (think SWG 18 or lower) with a high voltage, high current source? Or is this going to work against me, putting too much energy into the soil and surrounding material, effectively swamping the signal returns I'm looking for? Note that I'm talking about a very precisely controlled pulse or frequency, not just sticking a car battery across some fencing-wire coils! This could result in a much smaller TX coil, couldn't it?

advTHANKSance!
-PtB

We can distinguish targets by their TC and the response amplitude.
A small close target may give the same amplitude response than a large far away target, but the TC may be different.
For GB we want to select a specific target, the ground, to give the same response amplitude with the coil close to the ground or at a certain height, so that we can compensate for it.

We stack or integrate many results to get better S/N so it is not a matter of reducing the next pulse energy.

High magnetic field strength can produce a higher amplitude response, therefore good for S/N, if the other factors are not sacrificed for it.

[Davor]

It is also true that deeper objects are dynamically different than the shallow ones, and you can literally hear the difference, provided your Rx produces proportional tone, and not some binary abomination.

So shallow ones appear sharp, and deep ones appear soft, and it is also related to the coil geometry. You figure it out in the first 15 minutes in the field.

[bklein]

Pete,

http://www.miscel.dk/MiscEl/miscel.html

The miscel is a very useful calculator around metal detector design. I calculate my coils with it.
The TINKERER designs are not traditional PI designs. They are different animals, maybe ahead of their time.
The TINKERER coils are for the TINKERER designs only, not ideal for anything else.

Tinkerer

So using this calculator, what are the parameters you inputting and what type of coil are you using if you say intend to make an 8" mono.
It was choking on my inputs. Also it keeps saying the program is too old and to contact the author for a new version.