Conical coils

[Tinkerer]

I am always amazed of how much useful information comes together on this forum. Many thanks to all of you.
Rest asured that any valid information that I can come up with, will be posted here.
All that said, here is another question:
I am much intrigued about conical coils. Using the coil calculator from chemelec, (many thanks) I can see immediately how the hight of the cone reduces the amount of inductance dramatically. But what influence does the cone have on the reception of the eddy current? Or how does it influence the shape and intensity of the tx field?
Tinkerer

[ElectroNovice]

Hi Tinkerer,
I have never played round with conical coils so I can't answer your questions.
While coils are very important I think what is more important is flexability in the Pi platform you use.How the GP series of detectors really work is I suspect is only known by Candy. I personally doubt that this platform has the flexability to be optimized for larger,deeper targets(except by Candy).By flexability I mean the ability to change on the fly,pulse lenghts,pulse repetition rate,duty cycle,sample delay,sampling and averaging protocols,sample windows,circuit gain and dynamic range and the ability to critically damp the coil and change its resonant frequency , vary the Txand Rx bandwidth and improve the S/N particularly over high ferrite soils. These are where I believe the real gains will come from and then a high Q optimized coil is the icing on the cake.
ElectroNovice

[Tinkerer]

Flexible design

Hi EN,
I agree with you that a flexible design will help to get the most out of a given coil. However, if I aim for a detector that is capable to find very fine gold, I need a very short, (better below 10us) delay to the first sample. Only a coil with low capacitance will give that.
Then it needs to be a coil made of very thin wire, best Litz wire, so that the eddy currents of the coil itself, decay in less than 10us.
Then the coil has to be a very good antena, so that it picks up faint eddy currents from small gold.
Only if these parameters of the coil are met, will the flexible design have a chance to bring out the captured signals.
Please correct me if my thinking is wrong.
Tinkerer

[DOOLEY]

ref : conical coil

hello all, i to have been investigating the pros and cons of various coil designs,and it has lead me here,as well as many other places of no help,thanks google.

all below is "as far as i understand it" based on what ive learned , im no scientist or genius ,and could be all wrong, hope you see where im coming from.

the cone deisgn is a variation on the flat spiral or monofilar coil, but with a big differance.

been doing allot of studying theory over the last few weeks , and have come up with the conclusion that adjacent coil cancelation is whats limiting detection distance and sensitivity.

someone finds a way or an arangement to minimize or to stop this then were all sorted,untill then the laws of phisics as we know have us limited.

anyway, the flat spiral coil ,with a hole in the middle behaves much like a "normal coil" concentrating the magnetic field in the center ,and then returning back round to form a circle or compleate loop.

a flat spiral with no hole i have found is a bit weird, magnetic force flows accross the face and is concentrated in the center and is pretty weak on the outsides,read comments on people saying good for pin pointing and that sounds about right.

the cone however (big hole at the bottom) , with NO hole at the top ,say it has a divergance of 45 deg, when pulsed would be all either north or south on the inside ,dependant on current flow. same as a flat spiral but with 2 big differances.

with a flat spiral the flux moves over the surface towards the center or vise versa,
and the cone does the same but ,going to use the wrong word here, directs or aims it downwards, which is a good thing.

the other conclusion is that the magnetic field is uniform over the base due to the distance top to bottom, concentrated field in the center distance and all that.

sensitivity on detection should also be increaced as a returning signal would be hiting
the coil at the right angle to induce flux spin on every coil with little cancelation.

imagine 2 wires together with a space, 0 0
if a magnetic field hits from below _____^ right in the center both spin opposite so no voltage is created , well it is, but it canceled itself out.

___0
__0
_0
0
^^^^
now with the cone idea, all the coils are spun in the same direction when the magnetic field hits = voltage=detection.

00000000000000
^^^^^^^^^^
flat spiral above subjects coils to both directions of "spin" and the only coils that detect anything ar at either side as the ones in the middle cancle out, so they serve no purpose other than to look pretty.

i think about things, i experiment allot, ellectronics have been in me for 30 years,
when theres a gap in my understanding i HAVE TO clarify for my own sanity,like a dog with a new bone,i wont put it down.(the bone, not the dog!)

ANY THOUGHTS OUT THERE ???
HAVE I STUMBLED ONTO A NEW IDEA AND SHOULD HAVE PATENTED IT ??
USEUALY ,IF IVE THOUGHT OF IT , SOMEONE THOUGHT OF IT YEARS AGO.

if im right , then detector coils in the future are going to be one hell of a weird shape.

[Tinkerer]

Hi Dooley,

this thread is a few years old, but it has provoked not much interest at all.
I am glad that you have spent some time working on the idea.
I have myself not spent much time on it. Tried spiral wound coils with small holes and larger holes, so I have a bit of an idea how they work, but the conical coils are obviously much different.
It might be interesting to try a conical RX coil With a small hole in the center, together with a bundle wound TX coil. The results would prove or disprove your theory.

all the best

Tinkerer

[DOOLEY]

ref ; conical coils

yes , i see 2006 for the first post, it doesnt move fast on here does it ,
a conical coil at 45 deg seams to have the answers, since then i have seen a pic on the net of a conical coil "IN" another conical coil, which on the face of it looks as though it would make the intensity stronger , and also recieve sensitivity better.

which has lead me to draw the design below, interesting ??

a conical coil 12 " wide at the bottom at 45 deg would be 6 " high and people may laugh at your metal detector search head,,but...........

bellow the cone pointing out of the cone would only be 2.5 " high which is more acceptable,cross section of a circular coil

so i rekon this is a pretty good basic idea , size shape acceptable,
as far as performance goes appart from the bottom windings all the coils can only be spun one way so to speak when a field hits them, then only the other way with opposing field.

the round bits are the coils with one diameter sepperation to minimize intercoil cappacitance.
the square at the top to be removed, just made it easyer to draw.

just had a thought , this design could have as many "cells" as req.
here is just an inner ring and an outer ,could have another ring if larger surface area req.

power loss / emmision should be good , sensitivity would be hard to beat.

[Tinkerer]

Well, it is an interesting idea, only trying will tell if it will work.
What type of detector do you want to use with that?

Tinkerer

[DOOLEY]

conical coils

was thinking single coil pulse induction as tx should be on par with mono coil , but rx should be better .rekon this could be the answer to the pressent limit on depth for a given size of object.

wasn't worried about lenght of wire / resistance as more would be good for rx,
as far as tx goes , if amps isnt enough at 12v , will just drive it with 50 or 80v via voltage multiplier / blocking oscillator. built a few of those , got that down to a fine art.

last night was thinking of adding extra "sides" either flat or 45 deg up, to try to "shape field a bit more , at the very least it would extent the distance the inner polarity has to go before it can loop back round and return.

another crazy thought was , aluminium wire , less dense , lighter , poss a little more invisable on receive , ali wire still makes an electro magnet , hell power up a roll of aluminium shielding tape and it becomes a magnet.
ohms/foot a little higher due to lower conductivity than copper.

thought about carbon wire for total invisibility , but that stuff doent give off much , if any field as far as i know.

poss having a go with the idea today.

[AdrianM]

Hi DOOLEY,

imagine 2 wires together with a space, 0 0
if a magnetic field hits from below _____^ right in the center both spin opposite so no voltage is created , well it is, but it canceled itself out.

___0
__0
_0
0
^^^^
now with the cone idea, all the coils are spun in the same direction when the magnetic field hits = voltage=detection.

00000000000000
^^^^^^^^^^
flat spiral above subjects coils to both directions of "spin" and the only coils that detect anything ar at either side as the ones in the middle cancle out, so they serve no purpose other than to look pretty.

Where does the classic solenoid winding fit into this? Vertical field lines hit every conductor identically.

O
O
O
^

[DOOLEY]

0
0
0
0
^
on the face of it, no spin or very little , allthough force is unlikly to be spot on center so something will happen.
if the force hits anywhere to the right , or anywhere in the center of the "solenoid" then they will all spin one way , perfect ,
and outside the circle , for the same polarity of force , they will all spin the other way , but with a lower effect.
sensitivity is in the center as "both sides of the coil pick something up and are added together.
imagine magnet or force outside circle or coil , only one side picks it up.

i proberbly dont use correct terminology , and some physics master will shout at me and ruller my hand , but my terms , as a layman , make sense to me.

by my rekoning , an inline solenoid coil is near perfect as you can get , as long as you only want max sensitivity in the center , which we do , so all is well.
also quite low inter winding capacitance, but could be lower if coils have a gap , say , one wire thickness.
dont have fancy software or anything , just got a head and a notepad and pen.

if spaced to far appart , lower cappacity yes , but suspect less multiplied force ,
still , solenoid still more perfect than a random bundle with uneven cancelation here and there.

this is one of those conspiricy / pie in the sky guess's,
mankind has a long way to go in understanding everything , were learning and discovering all the time,
we think were advanced now , victorian people , with a steam train poss thought they were at the peak of technology,
when i was a kid in the 80's getting into electronics never imagined where we would be today.

as far as magnetism goes , in a hundred years time we will prob look back and laugh and say , "they were nearly there".

oh well , thats my 2 peneth.

[AdrianM]

I'm trying to answer entirely in your terms here (spin has, as you are aware, nothing to do with it really). I'm still wondering what makes you think that the precise centre of a conductor has such a critical role to play in the polarity of the EMF induced by a magnetic field passing to one side or the other.

[Tinkerer]

0
0
0
0
^
on the face of it, no spin or very little , allthough force is unlikly to be spot on center so something will happen.
if the force hits anywhere to the right , or anywhere in the center of the "solenoid" then they will all spin one way , perfect ,
and outside the circle , for the same polarity of force , they will all spin the other way , but with a lower effect.
sensitivity is in the center as "both sides of the coil pick something up and are added together.
imagine magnet or force outside circle or coil , only one side picks it up.

i proberbly dont use correct terminology , and some physics master will shout at me and ruller my hand , but my terms , as a layman , make sense to me.

For TX, a certain shape my be best, for RX a different shape may be best.

Magnetic fields can be pushed this way or that way. An example I like, is shown in my Avatar, where 2 rings conduct current in opposite direction. The field of the top coil pushes the field of the bottom coil down, so that there is a weak field on top and a stronger field at the bottom.
When 2 opposing fields deflect each other, then we get a concentration of field lines at the point where they meet. Similarly we will get a spot where the field lines are less concentrated because the overall amount of field lines does not change.

The more we pinch the field lines, the more they diverge when they are released, so a pinched field will penetrate less into the ground.

A bundled coil gives the greatest depth and a large diameter gives the greatest depth. However, with the increase of diameter, the field intensity tends to get weaker, so the sensitivity of the coil gets reduced.

For example:
We have a 300mm diameter, 300uH coil that we drive with 3A. The coil has 20 turns.
20x3=60
The coil has about 700cm square area therefore 60/700=about 0.085 field density.

If we take a coil of 600mm diameter and 300uH, the turns are about 13. With the same 3A drive we get 13x3=39
The 600mm coil has about 2800cm square 39/2800=about 0.014 field density, or 0.085/0.014= 6 times less field density.

Morale of the story, the field of the larger coil will penetrate deeper into the ground, but the field density will be less, so the size of the targets to be detected needs to be larger.

Now, how do we want to kick that target to get the best response?

Then, How do we want to configure the RX coil or coils, to best pick up the magnetic field emitted by the target?

And, how are we going to differentiate the response from the desired target from the response emitted by the surrounding matrix?

Tinkerer

[DOOLEY]

CONICAL COILS

adrianm , suspect i'm at the bottom of the pile for knollage on this subject ,
way i see it ,as previous clever scientist have said , electricity and magnatism are linked 100%
flow causes a magnetic field to "spin" round a wire at 90 deg to the wire yes ?
so a magnetic field will also cause a magnetic field around a wire to "spin" and so create flow , if same field is applied at both sides , there can be no "spin" so there can be no flow.
same as if force is centered on a wire , with equal force each side , there can be no "spin" and no flow.

the picture i have in my head is a wheel on an axle which is the wire,when a force is applied to the left right or center,
the wheel is either stationary ,going clockwise or going anti clockwise.

i got tierd of winding coils and trying them based on nothing concrete,
so looked into how it all happens , back to the basics.

it's only my understanding and interpritation , don't want to step on anyones toes or say that im talking fact , im just trying to work things out logicaly.

tinkerer , agree with you , one coil is a compromise , like jack of all trades , can do everything , but not perticulaly good at any of them.

sepperate coils has to be the way , one good for tx and the other good for rx.

never realy looked at your avitar till now , i would have thought one would cancel the other rather than direct or manipulate the field.

its a pitty dc magnatism cant be dirrected in the same way as we dirrect rf like yagi (tv) aerials , hell now that would sort it , focusing all energy into the ground , then only recieving from the ground with gain.hmmmmmmmmmm........am i dreaming there, may happen one day.

so anyway , i'm just thinking , we send a pulse , we want it to be as strong as poss to induce eddy currents in a target so that it can develop a magnetic field itself , and then we try to recieve as much of that as poss convert back into voltage / flow , without any noise from anywhere else,
so rather than putting up with a coil , the same coil or similar to what everyone else has and making do , just thought , can this be done better , have a look into it , how it all works , and try to do better.

[bklein]

.
.so rather than putting up with a coil , the same coil or similar to what everyone else has and making do , just thought , can this be done better , have a look into it , how it all works , and try to do better.

How about this:
Buy two surfPI boards and make them up. Test them on a coil to check for same performance. Then use one as your TX (with its own coil). Use the 2nd as your RX (with its own coil) - disable its TX function by taking out the Fet. You could bring over the TX sync signal to the RX board. Now you can play with RX or TX coils (types and placements) and document improvements. If you don't do it this way you are back to trying to the "minimize capacitance everywhere" battle, and it defocuses you from just nailing the basics of which type of coil is best for each purpose.

This is one way I intend to investigate this stuff. Another would be to make up a TX mapping system to better document TX coil/circuit performance - use hall-effect sensors or alternatives in 3D fashion. Wish I knew how to do software to have a cool realtime plot of results... Maybe I'll figure out a work around using available programs. But this is way off for me as I know I can't get the time currently. Be interested to learn from others if they've tried either suggestion.

Barry

[DOOLEY]

conical coils

bklein , yep , true comparison with a referance is the only way , could be done with only one board though , just cut the track from the fet drain to the resistor going to the diodes , then tag one side for rx , leaving the fet for tx.

incedentaly , tried this , and if tx and rx coils are close enough (ie:wound together) then you can choose weather to have damp resistor on tx and rx or just rx , i found that with sepperate but close coupled tx and rx coils the output waveform is "clean" with just a resistor on the rx coil,
it would apear that the tx can be as raw and dirty as you like. and as a bonus the rx resistor can be allot higher , i got a clen signal with 3.3k so less signal loss also.

oh , dont clamp the tx with diode back to batt + it kills your rx signal , tried it , just to see , no harm done , it just didnt help.

oh i also thought of hall effect sensors as an rx idea , 2 main types , north or south turn on types or just south turn on types , but there not realy that sensative , need alot of field to turn em on , think due to internal circuitry with output either on or off.

now if someone made a hall effect say 10" round and flat just on its own , no extra circuitry , ather than maybe a built in amp , with a variable signal out that was very sensative then we would have our rx sorted no ??

as far as software goes , i shy away from that , as soon as i see a project on the net with a pic chip in it i click the red X.

i dont want all that complication , its like , if you have a poor tv signal , if you amp it many times you useualy just get a stronger crap signal , if the signal ain't there , you cant amp it, thats why i thought back to basics , find a way to get a stronger signal right at the begining rather than try to get what you want with fancy chips and amping it to the point that it's unstable and temprimental.

best sensitivity i have had so far is a 12" coil with 30 turns of 20swg for tx , and then would on top 120 turns of 30 swg for rx , not built a full pi yet , just the tx side and the first amp , and scope'ing it.

benifit with seperate rx and tx is single supply is only needed as return signal is 0v and up , not 12v and up.
downside is , all that extra copper , knock sensitivity as its adding to capacitance and effectivly "sees" all that copper as a target.
went for a 30 to 120 ratio for a bit of "gain" right at the start.

so my next experiment is ali wire 0.8mm (welding wire actualy) with some .65mm(60lbs fishing line) nylon as a spacer to reduce capacitance , either "solenoid" wound or my 2 cone idea , dont know yet , but thinking as ali is less dense then would poss be more invisible as far as rx goes.

hoping for many people to join in on this till we get it cracked , hopefully before the big company do it before us , then want to charge a pretty penny for the pleasure.

last thing , not allot on the net regarding focusing magnetic force or similar , so is it not possible ?? not yet mastered ?? or "they" don't want us to know the secret ??