spiral pi coil

[bbsailor]

Theory tells us that to fully stimulate a target, the TX decay pulse time constant needs to be 5 times faster than the target time constant. The TX pulse TC for a 2us target would need to be 2 divided by 5 or .4uS. The TX decay TC is calculated by dividing the coil inductance by the effective damping resistor (Rd) value. bbsailor reply#92


Thinking about a larger coil. Coil I'm using now is Rx_(2)8inch round figure8, Tx surronds Rx. SRF(Tx and Rx) close to 1MHz. Should be good for targets above 1.6us TC. Thinking larger coil not good for faster TC targets, maybe .318 MHz SRF(1us TC)good for targets with TC greater than 5us might be a place to start. Is there a disadvantage to making the coil with the highest inductance other than weight?

Green,

I would optimize my coils diameter, inductance and lowest delay capability based on the targets I am primarily seeking. Getting deep coins requires a different design strategy than seeking low TC gold nuggets. Just reverse engineer the coil characteristics by defining the target TC, target size relative to coil diameter, and energy needed to stimulate the target at your desired depth. Low coil capacitance and higher coil self resonance is required for small low TC targets. The creativity comes from doing the balancing act to do the best optimization. Try to use shielding of the coil that adds the least capacitance. I sent Eric Foster a coil (300uH) that I made, when I wrote my Fast Coil article, with AWG30 Teflon insulated wire and a single layer of Scotch24 mesh shield and he said he could get it to operate at 5uS. If you are seeking small gold nuggets, operating at the lowest delay is very important and a coil size between 4 to 6 inches in diameter to optimize detection depth.

Build a lot of coils, try them with a wide range of targets at different depths and take notes. Eventually it will all come together.

I hope this helps.

Joseph J. Rogowski

[godigit1]

The flat coil calculator i suggested http://www.deepfriedneon.com/tesla_f_calcspiral.html calculates 298uH for AWG30, 28turns, 6.5inch ID. Maybe Q's calculator is closer?

I purchased a 20 x 30 x 3/16 inch foam board. Elmer's foam board #950805 at Walmart. Thinking of making a larger coil Rx_(2)round figure8, Tx surrounds Rx. Trying to determine coil size, wire size and number of turns for Rx and Tx. Rx on one side of board, Tx the other side.

Im thinking that the foam board was part of my low inductance problem with the spiral figure 8 coil. Im thinking the thickness between the crossover area is too thick. You would be wireing the coils together from side to side so it should be good for the eight portion. Im wondering if it will effect coupling between the Tx And Rx coils that one coil may be harder to balance.
Im not sure quite how that would effect the bandwidth and Q. would that give you a dual Bandwidth coil?
Its hard to find coil specific design literature so ive been trying to relate other fields as best I can.
Maybe you have seen these articles allready but if not there is some useful info.

http://frank.yueksel.org/other/RCA/R...-Inductors.pdf

https://digitalcommons.unl.edu/cgi/v...xt=physicsrudd.

I revisited this calculator and a couple other spiral coil calculators that I found. They were all the same as the tesla calc. Its strang though beacuse it hit the wire length very close At 50 feet , Im at about 54 with leads.
I played with the wire spacing, and for each per wire diameter it was giving me about 8uh reduction in inductance.
Im seeing quite a bit more than that on my actual coils.
For one wire space the inductance droped 50uh. on the coil.
Capacitance droped but not as much as it would with a different coil type.
Ill get specs up soon but so far without any target testing yet the wire spacing helps a little with adjusting the Q of the coil a bit.
I think it has a greater effect on different coil types like a bundle or stacked type of coil.


Good luck with your new design interested to see what you find out.

[green]

Calculated two coils with the flat spiral coil calculator. 23 inch diameter, 200 feet AWG24 wire, 1445uH and 23 inch diameter, 100 feet of wire, 355uH. The 355uH coil has higher signal and about 3 times higher peak coil current with spice simulation(same Tx on time). Signal increase less than 3 times. Maybe the 1445uH coil is better? Does anyone know if I made the two coils and tested them, would I get similar results? Wondering if I can use spice to minimize the number of coils I need to try.

[green]

Im thinking that the foam board was part of my low inductance problem with the spiral figure 8 coil. Im thinking the thickness between the crossover area is too thick. You would be wireing the coils together from side to side so it should be good for the eight portion. Im wondering if it will effect coupling between the Tx And Rx coils that one coil may be harder to balance.
Im not sure quite how that would effect the bandwidth and Q. would that give you a dual Bandwidth coil?
Its hard to find coil specific design literature so ive been trying to relate other fields as best I can.
Maybe you have seen these articles allready but if not there is some useful info.

Interested in what your learn with your coils. With my figure8 Rx coil the two round coils are connected in series. Rx inductance is the sum of two round coil inductances. Neither round coil is balanced with the Tx coil like a DD. To balance, either round coil is positioned to give the opposite signal of the other. How are you using bandwidth and Q to relate to PI coils? I'm thinking Tx resistance effects charge current profile. Coil + circuit capacitance effects Rd which controls discharge rate.

[godigit1]

" How are you using bandwidth and Q to relate to PI coils? I'm thinking Tx resistance effects charge current profile. Coil + circuit capacitance effects Rd which controls discharge rate."

Part of testing these coils has helped me to understand how to adjust SRF of a coil. So thats easy some might say. but what if you need a coil with a exact resistance and Srf range you have to engineer those perimeters out of a specific length of wire. Calculators could be played with for various wire diameret wrap and size combinations to get what we need. But capacitance and inductance are kinda locked in by those perimeters say if you are building a bundle coil as a example.
Lets look a Mr Rogowskis coil why is it faster than a regular bundle. The simple answer spacing added by the teflon insulation reduced capacitance between twists.
But what happened if im thinking correctly about this is .
Inductance was reduced. to get the same amount of inductance wire mass was increased.
Capacitance was lowered by quite a bit in this type of coil with some spacing. My initial test showed a 100 pf reduction with a similar coil.
Resulting in a higher SRF.
If we relate that to Q thats a hgher Q for the coil now.
Bandwidth was now narrowed and raised on the coil to make it more target specific or more sensitive to low conductors.

A lower Q is a higher bandwith so the same coil without the spacing added by the teflon coating makes a better coin hunter as it has wider bandwidth.

So there it is I hope Im even close but this is how it think it all relates to pi coils.
Im glad the coil gods that be are looking in from time to time. Thank you guys so much for your wisdom. Am I even Close?

I thought I would have the third coil with two wire spacing done last night but didnt get er done.
Specs for the first two are.

All coils are 30 awg enameled wire 54 feet with lead not twisted.
6.5 inch I.D flat spiral.

Coil1
No spacing between wraps.

5.9 ohms
390uh
1.389 MHZ Srf
33.664 Pf
Q 576.559

Coil two one wire spacing.
5.9 ohms
340 uh
30.469Pf
1.563Mhz
Q 565.10

So for coil 2 the Q droped just a little due to the drop in induction even though the Srf is higher. But they are very close I would call that a wash.
But the lower inductance and slightly lower capacitance I think will make it a bit faster. Not sure
Ive got to get a test Pi up and running to see any actual differances on targets, Better of worse .
For this coil design as close as capacitance and Srf are coil 1 I think will be the better because of a stronger field.
I think the 2 wire spacing coil will reduce values linerally like coil two should end up around 290 uh inductance.
I think better results will be seen building coils this way with the same inductances Ie changing wire Mass.

Im going to have to build one of your coils and try a small version of it. Thank you for explaining how you have it set up. I have to build one I dont understand how the Rx is not picking up the Tx.
Is this called direct coupling?
Thanks again.

[waltr]

That sounds good.
I can believe that for small low TC targets coil 2 would be better.

The simple answer spacing added by the Teflon insulation reduced capacitance between twists.

In addition, capacitance is also dependent of the dielectric constant of the material between the conductors. The lower the dielectric constant the less capacitance.
Teflon's dielectric constant is about 2.1 verse about 3 for PVC.
This would be seen in tests where conductor spacing is the same (same insulation thickness) and Only the wire insulation material changed.
How much faster this would make the coil is not known by me.
The articles using Teflon insulted wire to make a faster coil may have more to do with wire spacing than the dielectric constant of the insulation.

The Basket type coils employ spacing and geometry to reduce capacitance.

This thread is interesting and I will continue following.
Thanks for all the posts.

[green]

" How are you using bandwidth and Q to relate to PI coils? I'm thinking Tx resistance effects charge current profile. Coil + circuit capacitance effects Rd which controls discharge rate."

Part of testing these coils has helped me to understand how to adjust SRF of a coil. So thats easy some might say. but what if you need a coil with a exact resistance and Srf range you have to engineer those perimeters out of a specific length of wire. Calculators could be played with for various wire diameret wrap and size combinations to get what we need. But capacitance and inductance are kinda locked in by those perimeters say if you are building a bundle coil as a example.
Lets look a Mr Rogowskis coil why is it faster than a regular bundle. The simple answer spacing added by the teflon insulation reduced capacitance between twists.
But what happened if im thinking correctly about this is .
Inductance was reduced. to get the same amount of inductance wire mass was increased.
Capacitance was lowered by quite a bit in this type of coil with some spacing. My initial test showed a 100 pf reduction with a similar coil.
Resulting in a higher SRF.
If we relate that to Q thats a hgher Q for the coil now.
Bandwidth was now narrowed and raised on the coil to make it more target specific or more sensitive to low conductors.

A lower Q is a higher bandwith so the same coil without the spacing added by the teflon coating makes a better coin hunter as it has wider bandwidth.

So there it is I hope Im even close but this is how it think it all relates to pi coils.
Im glad the coil gods that be are looking in from time to time. Thank you guys so much for your wisdom. Am I even Close?

I thought I would have the third coil with two wire spacing done last night but didnt get er done.
Specs for the first two are.

All coils are 30 awg enameled wire 54 feet with lead not twisted.
6.5 inch I.D flat spiral.

Coil1
No spacing between wraps.

5.9 ohms
390uh
1.389 MHZ Srf
33.664 Pf
Q 576.559

Coil two one wire spacing.
5.9 ohms
340 uh
30.469Pf
1.563Mhz
Q 565.10

So for coil 2 the Q droped just a little due to the drop in induction even though the Srf is higher. But they are very close I would call that a wash.
But the lower inductance and slightly lower capacitance I think will make it a bit faster. Not sure
Ive got to get a test Pi up and running to see any actual differances on targets, Better of worse .
For this coil design as close as capacitance and Srf are coil 1 I think will be the better because of a stronger field.
I think the 2 wire spacing coil will reduce values linerally like coil two should end up around 290 uh inductance.
I think better results will be seen building coils this way with the same inductances Ie changing wire Mass.

Im going to have to build one of your coils and try a small version of it. Thank you for explaining how you have it set up. I have to build one I dont understand how the Rx is not picking up the Tx.
Is this called direct coupling?
Thanks again.

Thanks for the data and the explanation of Q. Wasn't relating Q to SRF. What procedure did you use to measure SRF? Your coil with no spacing has a lower SRF but a higher Q, which is better. Both are close enough to be the same. How fast does the coil need to be? Tested some 8 mesh nuggets(7/gram)awhile back. TC's varied between 1 and 2us. Smaller nuggets could be less than 1us. bbsailor suggests the coil TC should be 5 times less than the target TC. For a 1us TC target, coil TC should be .2us(SRF=1.59MHz). Assumes no added circuit capacitance, coil lead probably add at least 35pf. Tx circuit needs a MUR460 diode in series to isolate the mosfet capacitance. Need to add shielding. [Think target not detectable if coil TC is greater than target TC, Coil TC doesn't need to be 5 times less, probably should be at least 3 times]? Made the chart reply #109 with circuit SRF 1MHz, shouldn't be that hard to get circuit resonance of 1MHz. 1.59MHz could be difficult. If you add 35pf(lead capacitance) to your above coils, SRF would be close to 1MHz. The center chart data is for circuit resonance of 318kHz(1us TC). Is 318kHz good if not looking for small nuggets, very low TC targets? Probably not hard to get 318kHz.
If trying for 1MHz SRF. Is there any reason not to try for maximum inductance?

Some thoughts for discussion, maybe correct maybe not.

I'm using the figure8 Rx because of less noise pickup and reduces ground signal. Get better detection distance with my detector, don't know if I'm doing something wrong with mono coils. Disadvantage, need to make 3 coils to make one detector coil. Would be good to know how to design a coil for the application with mono coils and not have to build three coils for every try with my coil.

[nonkapo]

Hi Green,

Beacuse of the over use of foam and the coil only had the rigidity of a peice of paper the coil itself is not perfectly flat. It sits pretty flat but the surface has rolled a bit. the outer diameter is about 2 wire thicknesses lower than the I.d. it has maybe a 5 to 8 degree nice even tilt.
I need a better transformer for a foam cutting table I made. this way I can wind directly onto a cured formed slice of foam.

Just a suggestion and sharing of my experience with the spiral coils. To improve the rigidity i use sandwich of two coroplast boards, one coroplast and one foam board or sandwich of two foam boards. This way you need much less foam between and you can do shielding easier. I do not think the inductance problem in your case is caused by the foam as the measurements with the foam and without in my case hardly differ.

And one quick question. Where do you think is better to feed the spiral from to reduce the chance of detecting the feeder line? Inside the diameter or outside? I face a mechanical issue there as well, due to the shaft location.

[green]

How fast does the coil need to be? Tested some 8 mesh nuggets(7/gram)awhile back. TC's varied between 1 and 2us. Smaller nuggets could be less than 1us. bbsailor suggests the coil TC should be 5 times less than the target TC. For a 1us TC target, coil TC should be .2us(SRF=1.59MHz). Assumes no added circuit capacitance, coil lead probably add at least 35pf. Tx circuit needs a MUR460 diode in series to isolate the mosfet capacitance. Need to add shielding. [Think target not detectable if coil TC is greater than target TC, Coil TC doesn't need to be 5 times less, probably should be at least 3 times]? Made the chart reply #109 with circuit SRF 1MHz, shouldn't be that hard to get circuit resonance of 1MHz. 1.59MHz could be difficult. If you add 35pf(lead capacitance) to your above coils, SRF would be close to 1MHz. The center chart data is for circuit resonance of 318kHz(1us TC). Is 318kHz good if not looking for small nuggets, very low TC targets? Probably not hard to get 318kHz.
If trying for 1MHz SRF. Is there any reason not to try for maximum inductance?

Some thoughts for discussion, maybe correct maybe not.

I'm using the figure8 Rx because of less noise pickup and reduces ground signal. Get better detection distance with my detector, don't know if I'm doing something wrong with mono coils. Disadvantage, need to make 3 coils to make one detector coil. Would be good to know how to design a coil for the application with mono coils and not have to build three coils for every try with my coil.

Some questions and thoughts.

If we decide we need a circuit SRF of 1MHz(TC=.318us), should the amplifier have a frequency cutoff of 1MHz or would the MPP amplifier with a frequency cutoff of 300kHz be just as good for 1 to 2us TC targets?

If desired target TC's are greater than 3us, would it be better to design for SRF of 318kHz(1us TC) instead of 1MHz(TC .318us)? I have three other nuggets, 4grain(3.1s TC), 10grain(6.7us TC) and 18grain(5.7us TC)? Maybe can make a 318kHz coil that would detect the three nuggets better than a 1MHz coil? Are there desired targets with TC's less than 3us other than gold nuggets? Maybe gold chain and micro jewelry? The lower SRF would reduce the signal for some unwanted low TC targets. I've been trying to get the circuit as fast as I can, maybe for the challenge. Wondering how fast I should be trying for.

Should the amplifier frequency response be the same as the Rx coil frequency response? With an IB coil, does the Tx response need to be as fast as the Rx response?

[waltr]

Some questions and thoughts.

Should the amplifier frequency response be the same as the Rx coil frequency response? With an IB coil, does the Tx response need to be as fast as the Rx response?

The amplifier should have more bandwidth, rule of thumb is 10x for no distortion, than the coil delay else it will not accurately follow the decay curve.

The TX Turn-OFF response Must be fast. Else it will stretch out the decay curve and also not allow the target's Decay to produce the largest signal.

A little note about a past post in this thread:
The TX ON time needs to be about 5 times the target TC to fully excite the target. This implies that for small low TC targets a short TX ON pulse will do (25 us TX ON for a 5us target). This further implies, and Ferrous Toes has posted, that the TX rate (pps) can then be increased. This increased PPS also helps put a larger signal on the output of the integrator circuit output thus increasing sensitivity.

[green]

The amplifier should have more bandwidth, rule of thumb is 10x for no distortion, than the coil delay else it will not accurately follow the decay curve.

The TX Turn-OFF response Must be fast. Else it will stretch out the decay curve and also not allow the target's Decay to produce the largest signal.

A little note about a past post in this thread:
The TX ON time needs to be about 5 times the target TC to fully excite the target. This implies that for small low TC targets a short TX ON pulse will do (25 us TX ON for a 5us target). This further implies, and Ferrous Toes has posted, that the TX rate (pps) can then be increased. This increased PPS also helps put a larger signal on the output of the integrator circuit output thus increasing sensitivity.

Thanks for the reply. The TX ON time needs to be about 5 times the target TC to fully excite the target. Missed that statement in this thread, seen it in other threads. I've been referencing Typically, the energy in gold, being fully stimulated, falls to zero energy within the metal in 5 time constants. A 2uS target TC would be very hard to detect at 10uS (time delay) but at 8us could be detected. Theory tells us that to fully stimulate a target, the TX decay pulse time constant needs to be 5 times faster than the target time constant. The TX pulse TC for a 2us target would need to be 2 divided by 5 or .4uS. The TX decay TC is calculated by dividing the coil inductance by the effective damping resistor (Rd) value. A 300uH coil with a 750 ohm Rd is 0.4uS.reply #92.(SRF 795kHz)

The amplifier should have more bandwidth, rule of thumb is 10x for no distortion, than the coil delay else it will not accurately follow the decay curve. I'm guessing a PI doesn't need no distortion. bbsailor suggests the coil(circuit)time constant should be 5 times faster than the target time constant. For his example(2us target .4us coil)what does the amplifier time constant need to be? Is 5 times faster than the target enough, not enough or more than needed?

[waltr]

Thanks for the reply.

The amplifier should have more bandwidth, rule of thumb is 10x for no distortion, than the coil delay else it will not accurately follow the decay curve. I'm guessing a PI doesn't need no distortion. bbsailor suggests the coil(circuit)time constant should be 5 times faster than the target time constant. For his example(2us target .4us coil)what does the amplifier time constant need to be? Is 5 times faster than the target enough, not enough or more than needed?

For a PI detector amplified BW of 5x coil decay should be good. Note: I said 10x for no distortion. We can allow some distortion but not too much since it changes the decay curve.

[green]

For a PI detector amplified BW of 5x coil decay should be good. Note: I said 10x for no distortion. We can allow some distortion but not too much since it changes the decay curve.

Tried a circuit in spice. C2 sets amplifier TC, 17.1pf for amplifier TC=coil TC=1/5 target TC. Change C2 to change amplifier TC, 85p for amplifier TC=target TC. Don't know if simulation would match real circuit or if the simulation is correct. Amplifier TC=coil TC looks good. Changing C2 to 34pf doesn't look to bad.

[godigit1]

Thanks for the data and the explanation of Q. Wasn't relating Q to SRF. What procedure did you use to measure SRF? Your coil with no spacing has a lower SRF but a higher Q, which is better. Both are close enough to be the same. How fast does the coil need to be? Tested some 8 mesh nuggets(7/gram)awhile back. TC's varied between 1 and 2us. Smaller nuggets could be less than 1us. bbsailor suggests the coil TC should be 5 times less than the target TC. For a 1us TC target, coil TC should be .2us(SRF=1.59MHz). Assumes no added circuit capacitance, coil lead probably add at least 35pf. Tx circuit needs a MUR460 diode in series to isolate the mosfet capacitance. Need to add shielding. [Think target not detectable if coil TC is greater than target TC, Coil TC doesn't need to be 5 times less, probably should be at least 3 times]? Made the chart reply #109 with circuit SRF 1MHz, shouldn't be that hard to get circuit resonance of 1MHz. 1.59MHz could be difficult. If you add 35pf(lead capacitance) to your above coils, SRF would be close to 1MHz. The center chart data is for circuit resonance of 318kHz(1us TC). Is 318kHz good if not looking for small nuggets, very low TC targets? Probably not hard to get 318kHz.
If trying for 1MHz SRF. Is there any reason not to try for maximum inductance?

Some thoughts for discussion, maybe correct maybe not.

I'm using the figure8 Rx because of less noise pickup and reduces ground signal. Get better detection distance with my detector, don't know if I'm doing something wrong with mono coils. Disadvantage, need to make 3 coils to make one detector coil. Would be good to know how to design a coil for the application with mono coils and not have to build three coils for every try with my coil.

Sorry for just getting back had a bad couple days heads still not quite on straight.

Ive been using Mpp rev d to get my Srf readings. Ive been taping the coils to the bottom of a 12 inch dual field coil.
your right both coils came out very close and the third one is not much different Im a bit suprised at how the values sorta wash out. Thinking that the wire resistance has something to do with it. Im not totally sure wich is better as far a Q goes Im wondering if I dont have things mixed up as far as target Tc To Bandwidth.
Im still trying to figure all this out but you guys have been really leading things in the right direction. Thanks you all I keep saying that And I mean it.

How fast does the coil need to be?
I was starting to mod the Mpp Pre amp In Telenos thread "Mpp Preamp is vastly improvable"
His disign is possible that it can get it down around 3us. For test purposes I would like a coil that fast.

I think as far as being used for hunting the ability to go to 5 us in the right ground woud be ok.
So how fast does a coil have to be in order to be a five us coil after the shielding and cable are added if using a cable.
Im still trying to wrap my head around the Chart but it looks like you are saying I have About 35 pf to play with for shielding? As all the coils allready have leads.
That would make the coil capable of .2 us TC ?


" Is there any reason not to try for maximum inductance?"

Good question. I will be curious to see how the three different coils act. Really inductance was the perimeter adjusted the most. Capacitance very little, Surely not enough to justify the extra effort of two line width spacing.
Here are the specs for the first two again to compare with three.

Coil1
No spacing between wraps.

5.9 ohms
390uh
1.389 MHZ Srf
33.664 Pf
Q 576.559

Coil two one wire spacing.
5.9 ohms
340 uh
30.469Pf
1.563Mhz
Q 565.10


Third coil two wire spacing.
5.9 ohms
298uh
1.667Mhz
30.588pf
Q530.5

They are all so close so back to your question why not go for max inductance in a case like this anyway?

Ive been procrastinating on getting a test unit up it involves breaking down other projects right now. having a hard time deciding the victim. Figured might as well get a few coils done in the mean time .

Have you tried a figure 8 mono with your Detector?

[green]

Hi godigit, Interesting tests. I haven't tried a figure 8 mono, just the figure 8 Rx with oval Tx. Very small nuggets are one of your desired targets? I'm interested in how short a delay you can get the MPP. I've been thinking amplifier TC should be close to coil TC. MPP has a 1us TC?