IGSL

[simonbaker]

I will re-post it here for you, both parts are very interesting and I also feel they belong to IGSL people as much as TGSL....its from Don.

I hadn't read part II until now - thanks again Don for great instructional!

Oops, my clevis is showing...

-SB

[der_fisherman]

An idea to help VLF Performance with Tesla Coils

I wanted to explain why I am building a semi „flat“coil for my IGSL with a TGSL (from TGSL 101 by Don Bowers) coil design with a 25cm diameter.

The idea came to me after reading up on Tesla flat coils that many people believe are particularly good for PI metal detectors. There was an explanation on one website (I have forgotten which so don’t ask!) about how the normal coils we make for our detectors, (round in section with the wires not following any particular path – e.g. jumbled), the wires do not each get an even receive effect from the received signal, some losses occur, efficiency is down as in a Tesla coil, each winding has an almost unobstructed “view” of the transmitted signal. No other windings “in the way” reducing the signal strength…...also the received signal is in “phase” over the whole of the coil, whereas a “jumbled” coil will receive a jumbled signal where some will cancel out others….

Also, I understand that “inter-winding” capacitance is also reduced, which also improves the ability of the coil to receive better……

I found a website that describes various coil types here:-
http://en.wikipedia.org/wiki/Coil
or here for flat Tesla coils:-
http://en.wikipedia.org/wiki/Tesla_coil
and there is a simple picture of a Tesla coil that I will attach to the post.

Quoting from the 2nd link I found this:-
As the primary energy transfers to the secondary, the secondary's output voltage increases until all of the available primary energy has been transferred to the secondary (less losses). Even with significant spark gap losses, a well-designed Tesla coil can transfer over 85% of the energy initially stored in the primary capacitor to the secondary circuit. Thus the voltage gain of a Tesla coil can be significantly greater than a conventional transformer, since the air gap has a very high insulation.

Now I am (sadly) unable to think of a fully practical procedure to make a totally flat coil with 0.25mm copper wire, in the required “D” shape for a VLF Metal detector, I think that I will be able to make quite a thin one, with approx. a 4 windings thickness, by using a CD as a spacer. When I find a thinner spacer one day, I will try and make one only 2 windings thick!! (But I may actually be taking on more than I can do even when it’s only 4 wires thick!!!! Don’t laugh if I fail please….?) But that is all for the future….in the coming 7 days I will be making the 4 wire thickness version….

4 wires thick (about 1mm) will make the coil’s wound width approx. 2.6 mm wide and that will let me make a 25 cm wide fully finished Double “D” search head with fairly flat coils……and expose directly more windings to the received signal than with normal jumbled wound coils, or so I believe……we will see….

The best way to make an even thinner coil x 2 would be to make them from PCB material I feel, as thin as possible, and mount them both on say plywood for strength, soaked in Epoxy for waterproofing, insulation and strength…..but I have not yet tried that myself either…..maybe never!!!
There is a picture of a Tesla coil from Gerry’s website here:-
http://www3.telus.net/chemelec/Projects/Coils/Coils.htm

The reason for posting this idea is to simply provoke thoughts from some of the good minds here…..so please fire up you best typing fingers and answer/query my thoughts.

Thanks in advance.

Andy

PS. Even if it is not better than a normal coil, it should not be any worse either!!!! Provided I null it out well of course. I will use Don B's setup procedures from TGSL 101.

PPS. You can call me crazy if you wish, I might even agree with you if the coil proves to be a disaster!!!

PPPS. As a double D coil search head, the coils are laid one over the other, one reversed, must I wind one of the coils "backwards" to the other to achieve optimum results, as making both on the same former and reversing one means that the coil will actually be the other way round (which does mean that the induced voltage is traveling through the overlaps in the same direction.....which may be a good idea.....

.......I know I need to know the beginnings and the ends, but nobody mentioned my question....maybe it makes no difference. If the phasing is wrong, I can reverse the connections to one coil.....but it would be nice to know. any ideas? thoughts? In the second picture, should the Rx signal be running in the same direction as with the two blue arrors, or should it be running against as shown with the red arrow?

PPPPS. It is recommended to place the Rx coil at the bottom to receive better, without the Tx coil in the way, so to say......any ideas? thoughts?

[simonbaker]

I wanted to explain why I am building a semi „flat“coil for my IGSL with a TGSL (from TGSL 101 by Don Bowers) coil design with a 25cm diameter.

The idea came to me after reading up on Tesla flat coils that many people believe are particularly good for PI metal detectors. There was an explanation on one website (I have forgotten which so don’t ask!) about how the normal coils we make for our detectors, (round in section with the wires not following any particular path – e.g. jumbled), the wires do not each get an even receive effect from the received signal, some losses occur, efficiency is down as in a Tesla coil, each winding has an almost unobstructed “view” of the transmitted signal. No other windings “in the way” reducing the signal strength…...also the received signal is in “phase” over the whole of the coil, whereas a “jumbled” coil will receive a jumbled signal where some will cancel out others….

Also, I understand that “inter-winding” capacitance is also reduced, which also improves the ability of the coil to receive better……

I found a website that describes various coil types here:-
http://en.wikipedia.org/wiki/Coil
or here for flat Tesla coils:-
http://en.wikipedia.org/wiki/Tesla_coil
and there is a simple picture of a Tesla coil that I will attach to the post.

Quoting from the 2nd link I found this:-
As the primary energy transfers to the secondary, the secondary's output voltage increases until all of the available primary energy has been transferred to the secondary (less losses). Even with significant spark gap losses, a well-designed Tesla coil can transfer over 85% of the energy initially stored in the primary capacitor to the secondary circuit. Thus the voltage gain of a Tesla coil can be significantly greater than a conventional transformer, since the air gap has a very high insulation.

Now I am (sadly) unable to think of a fully practical procedure to make a totally flat coil with 0.25mm copper wire, in the required “D” shape for a VLF Metal detector, I think that I will be able to make quite a thin one, with approx. a 4 windings thickness, by using a CD as a spacer. When I find a thinner spacer one day, I will try and make one only 2 windings thick!! (But I may actually be taking on more than I can do even when it’s only 4 wires thick!!!! Don’t laugh if I fail please….?) But that is all for the future….in the coming 7 days I will be making the 4 wire thickness version….

4 wires thick (about 1mm) will make the coil’s wound width approx. 2.6 mm wide and that will let me make a 25 cm wide fully finished Double “D” search head with fairly flat coils……and expose directly more windings to the received signal than with normal jumbled wound coils, or so I believe……we will see….

The best way to make an even thinner coil x 2 would be to make them from PCB material I feel, as thin as possible, and mount them both on say plywood for strength, soaked in Epoxy for waterproofing, insulation and strength…..but I have not yet tried that myself either…..maybe never!!!
There is a picture of a Tesla coil from Gerry’s website here:-
http://www3.telus.net/chemelec/Projects/Coils/Coils.htm

The reason for posting this idea is to simply provoke thoughts from some of the good minds here…..so please fire up you best typing fingers and answer/query my thoughts.

Thanks in advance.

Andy

PS. Even if it is not better than a normal coil, it should not be any worse either!!!! Provided I null it out well of course. I will use Don B's setup procedures from TGSL 101.

PPS. You can call me crazy if you wish, I might even agree with you if the coil proves to be a disaster!!!

PPPS. As a double D coil search head, the coils are laid one over the other, one reversed, must I wind one of the coils "backwards" to the other to achieve optimum results, as making both on the same former and reversing one means that the coil will actually be the other way round (which does mean that the induced voltage is traveling through the overlaps in the same direction.....which may be a good idea.....

.......I know I need to know the beginnings and the ends, but nobody mentioned my question....maybe it makes no difference. If the phasing is wrong, I can reverse the connections to one coil.....but it would be nice to know. any ideas? thoughts? In the second picture, should the Rx signal be running in the same direction as with the two blue arrors, or should it be running against as shown with the red arrow?

PPPPS. It is recommended to place the Rx coil at the bottom to receive better, without the Tx coil in the way, so to say......any ideas? thoughts?

Always good to experiment - we'll be interested in what you find.

My intuition says that a "full spiral" coil as depicted in the picture would not be better for metal detecting -- because the magnetic field outside a loop is the opposite direction as the field inside the loop -- so the exterior field from the smaller loops will oppose the interior field of the larger loops. At least at the surface of the coil.

Deeper below the coil, I would think the field from the smaller loops simply won't penetrate deep at all, so the field will be mainly from the larger loops -- and there are fewer large loops than with a conventional coil where all loops are large; i.e. weaker field.

However, if you made a "partial spiral" design where the windings are spiral but all are fairly large loops so it is still similar to a conventional coil, that might work better. If the wire is thin enough, it might work out. Your "2.6 mm flat coil" idea sounds good for that.

Your point about loops shielding each other (in a conventional coil) -- I'm not sure that really happens. If it does, it would also apply to the TX coil as well. Maybe there is a small effect. Need to break out the math or some simulation program.

I think inter-winding capacitance may not matter for VLF MDs like the TGSL/IGSL, because we put a big tuning capacitor across the coil anyway and the interwinding capacitance may just effectively add to it a small amount.

-----------------------

As to how the wires are oriented (direction of winding), I believe it doesn't matter; all that matters is the direction of the current, and you can control that by just reversing the leads if you get it wrong (before you pot the coil if you use the "grounded RX coil" design! -- because you want the grounded lead to go to the correct op amp input).

As to which coil is on top -- I suspect you wouldn't be able to measure a difference. There may even be "reciprocity" theory where it literally works out the same either way. Even from a signal-to-noise ratio consideration it may not make a difference. But I put the RX coil on the bottom anyway!

Just my thoughts. I think it is great to experiment and let us know!

-SB

P.S. Maybe we should pursue this in a different thread to reduce confusion here. Qiaozhi can relocate it if desired.

[ivconic]

Thanks guys!
I will read it and see.

[dfbowers]

Always good to experiment - we'll be interested in what you find.

My intuition says that a "full spiral" coil as depicted in the picture would not be better for metal detecting -- because the magnetic field outside a loop is the opposite direction as the field inside the loop -- so the exterior field from the smaller loops will oppose the interior field of the larger loops. At least at the surface of the coil.

Deeper below the coil, I would think the field from the smaller loops simply won't penetrate deep at all, so the field will be mainly from the larger loops -- and there are fewer large loops than with a conventional coil where all loops are large; i.e. weaker field.

However, if you made a "partial spiral" design where the windings are spiral but all are fairly large loops so it is still similar to a conventional coil, that might work better. If the wire is thin enough, it might work out. Your "2.6 mm flat coil" idea sounds good for that.

Your point about loops shielding each other (in a conventional coil) -- I'm not sure that really happens. If it does, it would also apply to the TX coil as well. Maybe there is a small effect. Need to break out the math or some simulation program.

I think inter-winding capacitance may not matter for VLF MDs like the TGSL/IGSL, because we put a big tuning capacitor across the coil anyway and the interwinding capacitance may just effectively add to it a small amount.

-----------------------

As to how the wires are oriented (direction of winding), I believe it doesn't matter; all that matters is the direction of the current, and you can control that by just reversing the leads if you get it wrong (before you pot the coil if you use the "grounded RX coil" design! -- because you want the grounded lead to go to the correct op amp input).

As to which coil is on top -- I suspect you wouldn't be able to measure a difference. There may even be "reciprocity" theory where it literally works out the same either way. Even from a signal-to-noise ratio consideration it may not make a difference. But I put the RX coil on the bottom anyway!

Just my thoughts. I think it is great to experiment and let us know!

-SB

P.S. Maybe we should pursue this in a different thread to reduce confusion here. Qiaozhi can relocate it if desired.

Just when I thought I read everything in the forum that was of interest, I find this:
http://www.geotech1.com/forums/showthread.php?t=13858

[ivconic]

I will re-post it here for you, both parts are very interesting and I also feel they belong to IGSL people as much as TGSL....its from Don.

I just finished reading those articles!
Wow!
Excellent job Don Bowers!
Pretty literate and illustrative!
Those articles are real reference!
Everything i agree, no significant remarks!
Maybe only "USB" cable use.. but that's already explained.
I am also taking 6mH and 6.5mH as inductances when making my own coil...but fact is that
original Tesoro coils are 6.2mH and 5.7mH, mostly!
The other day i had Conquistador on service.. and measured also inductances... 6.2 and 5.7...
Same as on Bandido i had few years ago... same as on Eldorado ...
Not so important... because 6.5 and 6 will also work nicely..

[dfbowers]

I hadn't read part II until now - thanks again Don for great instructional!

Oops, my clevis is showing...

-SB

I'm glad that some can make use of it as it makes me feel that I didn't waste a weeks worth of evenings. I still think that section#1 is lacking some detail that I believe is important. I had to limit the content to 2MB per section. Still need the troubleshooting section though!

Thanks Simon.

[ivconic]

I'm glad that some can make use of it as it makes me feel that I didn't waste a weeks worth of evenings. I still think that section#1 is lacking some detail that I believe is important. I had to limit the content to 2MB per section. Still need the troubleshooting section though!

Thanks Simon.

By no means! It is not wasted time at all!
I will print those documents right now and will have them in my workshop!
Excellent done!

[ivconic]

Maybe you don't know; but i've been absent from Geotech for longer period. I haven't followed TGSL related topics since than.
I did quite job on TGSL in the past.. and once when i was tired of it - i simply left others to continue.

I don't know about Don' Bowers works nor i don't know about any "wet Grass"...whatever it may be!?

Sheeeeeeeeeeeeeeeeeeshhhhh!

Don Bowers is Dfbowers!!!!!
I must be gone senile!!
It took me whole day to dug this!!!
Of course i DO KNOW about Don Bowers (as DFBOWERS) !!!!
But fast reading and viewing posts completely confused me and i took the name pretty "automatically" !!!

[ivconic]

Done, printed!
Those will be perfect reminder!
Usually i do keep my notes spread on million small pieces of paper....
Worth nothing, because i can't find them exactly when i need them!
Excellent done!

[Jerry]

I'm glad that some can make use of it as it makes me feel that I didn't waste a weeks worth of evenings. I still think that section#1 is lacking some detail that I believe is important. I had to limit the content to 2MB per section. Still need the troubleshooting section though!

Thanks Simon.

It was time well spent Don! I printed mine out and put in it's own binder but forgot where the original post was located so had to do a search in order to find it again. Perhaps the administrators can make this a sticky along with the complete details one.

Jerry

[ivconic]

It was time well spent Don! I printed mine out and put in it's own binder but forgot where the original post was located so had to do a search in order to find it again. Perhaps the administrators can make this a sticky along with the complete details one.

Jerry

That's good idea!
I vote YES!

[der_fisherman]

Tesla Coils for IGSL Detectors

Firstly, many thanks Simon, you gave me a lot to think about (exactly what I wanted by the way!) I hope I get some more replies similar to yours, thanks. I thought that I would answer each (excellent) point that you made:-

My intuition says that a "full spiral" coil as depicted in the picture would not be better for metal detecting -- because the magnetic field outside a loop is the opposite direction as the field inside the loop -- so the exterior field from the smaller loops will oppose the interior field of the larger loops. At least at the surface of the coil.

I am open on this point, but this happens already with ANY coil, in fact, if I could ever draw in Eagle CAD AND it was the correct value of impedance (unlikely), I would love to try 2 of them on an IGSL in double D format of course.....That might be an awesome thing to make and use.......untill I see one, I remain unconvinced that it will not be far better than a conventional coil......that's just me, I could easily be 100% wrong.....

Deeper below the coil, I would think the field from the smaller loops simply won't penetrate deep at all, so the field will be mainly from the larger loops -- and there are fewer large loops than with a conventional coil where all loops are large; i.e. weaker field.My thoughts are that this happens right now with any coil anyway

What I believe is similar to what you say, but that the Tesla coil will work better (at the same time) for both large and small objects as it is both a large and a small coil (and a medium one too!!) and we both know that with a larger coil you get more depth and see large objects better and with a small coil you get less depth, but more sensitive to smaller objects. Now you can have both at the same time (I really do believe this!).

However, if you made a "partial spiral" design where the windings are spiral but all are fairly large loops so it is still similar to a conventional coil, that might work better. If the wire is thin enough, it might work out. Your "2.6 mm flat coil" idea sounds good for that.

Agreed. But don't forget that a true flat TGSL Rx coil(107 turns) "a la Tesla", assuming that the outside radius would be for a 25cm search head, made from 0.25mm copper wire (actually 0.26mm with coating) would be only a band of 2.782 cm wide or about 1.095". It would not actually look much like the picture I posted from Wiki.....The hole in the middle would be almost 20 cm across.......so its appearance would be relatively similar in many ways to a conventional coil.....7.6" inner hole and 9.84" outside diameter - but flat!

Your point about loops shielding each other (in a conventional coil) -- I'm not sure that really happens. If it does, it would also apply to the TX coil as well. Maybe there is a small effect. Need to break out the math or some simulation program.

I seriously doubt whether the effect of shielding the Tx coil, which is radiating almost 100% of the energy used in the Collpits Oscillator, has as large an effect as shielding the Rx coil which is receiving maybe 1% of that energy or less......at best. Remember the Rx coil sits in the null and what we eventually pick up has to be seriously amplified before we can use it.....but it does demonstrate to me at least, that windings shield windings according to you as well. So a flat Tesla Tx coil could even improve Tx radiation as well.....because its not "blocked" anywhere.....(Except where the Rx coil overlaps, underneath of course!!)

I think inter-winding capacitance may not matter for VLF MDs like the TGSL/IGSL, because we put a big tuning capacitor across the coil anyway and the interwinding capacitance may just effectively add to it a small amount.

I agree. Though I have actually no idea what the actual value of capacitance from say a TGSL Rx coil actually is.....guessing only a few pFs....but as once my IGSL is finished, I am going to build a Russian PI machine as well, I am very interested in making a good job for that too.......

As to how the wires are oriented (direction of winding), I believe it doesn't matter; all that matters is the direction of the current, and you can control that by just reversing the leads if you get it wrong (before you pot the coil if you use the "grounded RX coil" design! -- because you want the grounded lead to go to the correct op amp input).

You and I have the same thoughts on that, if the phasing is wrong, just swap the connections to one coil only. By the way, a good reason to bring the coil and shield connections out in the end of the cable and not to connect them in the search head - ever!!!

As to which coil is on top -- I suspect you wouldn't be able to measure a difference. There may even be "reciprocity" theory where it literally works out the same either way. Even from a signal-to-noise ratio consideration it may not make a difference. But I put the RX coil on the bottom anyway!

LOL - but you put it on the bottom anyway, "just in case?" Simper, simper!

P.S. Maybe we should pursue this in a different thread to reduce confusion here. Qiaozhi can relocate it if desired.

I do agree that this post it could be out into Coil making, but really it also needs to be here for IGSL construction, once we are all agreed on what is a good method....which may be a week or two at least, I am a slow builder at the best of times!!!

Simon, many thanks for your very active and keen interest, I have the impression that little of this was really new for you, I would guess that you have also been having some thoughts in this area. I suspect we are not alone in this matter either!!!

Regards

Andy

[der_fisherman]

Addendum for TGSL 101?

How about an "Addendum" for part 1 if you have some extra info's that did not make it into the original document?

I for one would be VERY interested to read such thoughts and comments.

By the way, I have never written a document as Author (I translate books/documents in other languages into English ) for the computer firm I worked for, and not had to add details months later, its perfectly normal.....

Its good to see that some others are also now fans of "TGSL 101", including Ivconic!!! You did a good and useful job with the original version as well!! In reality, I expect it has been read by 100s of members here....if not even more.

Regards

Andy

[ivconic]

" Originally Posted by simonbaker
As to how the wires are oriented (direction of winding), I believe it doesn't matter; all that matters is the direction of the current, and you can control that by just reversing the leads if you get it wrong (before you pot the coil if you use the "grounded RX coil" design! -- because you want the grounded lead to go to the correct op amp input)."

" Reply from Andy
You and I have the same thoughts on that, if the phasing is wrong, just swap the connections to one coil only. By the way, a good reason to bring the coil and shield connections out in the end of the cable and not to connect them in the search head - ever!!!"

This is most interesting part for me.
Would be good to clear this up.
Yes, you will achieve (virtually) same effect ...but is it the same for real?
I don't have proper answer. To answer that; i think man must have more advanced measuring equipment...than i have..
I wander what Qiaozhi, Aziz... and others can say on this subject?