Optimizing PI Coils

[green]

Been thinking making another coil to better detect #9 lead shot. The one I'm using. (Rx)two round 133mm coils side by side connected figure8 opposing, (Tx)oval coil over Rx. Wound with AWG28 magnet wire. 133mm is inside measurement of each spiral wrapped coil. Thinking of trying 100mm figure8 using AWG30 magnet wire. Single winding for each Rx, two stacked windings for oval Tx. Thoughts or anything else I might consider.

[Carl-NC]

What do you mean "two stacked windings for TX"?

[green]

What do you mean "two stacked windings for TX"?

Wind a single layer spiral coil, then wind another on top of it. Connect in parallel, 2 AWG30 in parallel same as one AWG27. Less resistance than my one layer AWG28. Winding the two layers at the same time would be better but I don't think I can without making something.
AWG28 has a time constant of .31us, AWG30 .19us. Hoping I can sample a little sooner.

Another option. Twist two wires together and wind a single layer spiral coil. Would make a wider coil. Any thoughts which might be better?

[bbsailor]

Wind a single layer spiral coil, then wind another on top of it. Connect in parallel, 2 AWG30 in parallel same as one AWG27. Less resistance than my one layer AWG28. Winding the two layers at the same time would be better but I don't think I can without making something.
AWG28 has a time constant of .31us, AWG30 .19us. Hoping I can sample a little sooner.

Green,

Try reshaping your round coils to oblong coils with a parallel straight section and a semicircle on each end. As you make the shape wider the area of the coil will be reduced and the inductance will be reduced. Try to find the optimum shape for detecting your desired target. As Carl said the #9 lead shot may best be detected near the coil edge. With an oblong coil about .5 to 1 inch wide parallel section you will be closer to the coil edge and also cover more ground in a single sweep. To optimize this coil shape you would need to make some different prototype shapes and test them with your #9 lead shot. Just optimize the coil inductance with a few more turns as the coil gets thinner to keep the inductance the same. Optimize the damping resistance value for each shape you try.

I recall Eric Foster making a long oblong coil to detect broken needles in spools of fabric. These broken needles had a very low time constant and passed very close to the long coil spanning the fabric width. You may be able to optimize this concept to detect your #9 lead shot.

I hope this gives you something to ponder and possibly try.

Joseph J. Rogowski

[green]

Green,

Try reshaping your round coils to oblong coils with a parallel straight section and a semicircle on each end. As you make the shape wider the area of the coil will be reduced and the inductance will be reduced. Try to find the optimum shape for detecting your desired target. As Carl said the #9 lead shot may best be detected near the coil edge. With an oblong coil about .5 to 1 inch wide parallel section you will be closer to the coil edge and also cover more ground in a single sweep. To optimize this coil shape you would need to make some different prototype shapes and test them with your #9 lead shot. Just optimize the coil inductance with a few more turns as the coil gets thinner to keep the inductance the same. Optimize the damping resistance value for each shape you try.

I recall Eric Foster making a long oblong coil to detect broken needles in spools of fabric. These broken needles had a very low time constant and passed very close to the long coil spanning the fabric width. You may be able to optimize this concept to detect your #9 lead shot.

I hope this gives you something to ponder and possibly try.

Joseph J. Rogowski

The two round coil figure8(top picture)charts close to the same distance vs signal amplitude as a mono with the same diameter as one of the round figure8 coils. Might be something to try. Any thoughts for the first attempt? No over lap(bottom left), more or less overlap(bottom right)? Has anyone tried it? Any thoughts on reduction of signal vs distance, what diameter to try.

Picture of Rx connected opposing, Tx round or oval over Rx. Could try with AWG28 to compare with the 2 round coil figure8.

[Carl-NC]

Another option. Twist two wires together and wind a single layer spiral coil. Would make a wider coil. Any thoughts which might be better?

I would expect the twisted wires to be worse. Twisted wires will be longer and have increased resistance and capacitance.

[green]

Don't know how to calculate required number of turns for odd shape coils, not round for same inductance. Thinking of winding some coils, 150mm inside diameter with 15 turns. Wind the mirror image and connect figure8(Rx). Wind oval Tx over Rx. Chart signal amplitude vs target distance at location a, b, c, d, e and f corrected for 15 ampere turns to see which coil looks better. I'm guessing the round Rx coils. If anyone knows it would save me some time.

[bbsailor]

Green,

The web site 66pacific.com has many coil size calculators. Round coils have the most area for a given circumference but if you squeeze the round coil into an oval the inductance will fall relative to the less area of the oval coil with the same coil perimeter length. The area of an oval squeezed to 1/2 the coil diameter will have its area drop to 75 percent of the round coil. If you squeeze the coil to 1/4 of the coil diameter, the area will drop to 62.5 percent of the circle area. Use an inductance meter on the round coil and when you move the coil into an oval, watch the inductance drop.

What depth are you seeking for your #9 lead shot?

Joseph J. Rogowski

[green]

Green,

The web site 66pacific.com has many coil size calculators. Round coils have the most area for a given circumference but if you squeeze the round coil into an oval the inductance will fall relative to the less area of the oval coil with the same coil perimeter length. The area of an oval squeezed to 1/2 the coil diameter will have its area drop to 75 percent of the round coil. If you squeeze the coil to 1/4 of the coil diameter, the area will drop to 62.5 percent of the circle area. Use an inductance meter on the round coil and when you move the coil into an oval, watch the inductance drop.

What depth are you seeking for your #9 lead shot?

Joseph J. Rogowski

Was thinking if I wind them the same number of turns I could calculate required turns for desired inductance once I know which coil shape is best. 150mm diameter round spiral calculates(79uH with 15T). Tx being oval should be more than 80uH.


I'm looking for the best coil shape and diameter to get 1 to 2 inch detection for #9 lead shot. Maybe if I get smarter that might change. First is defining shape, diameter, turns and wire size. Maybe others.

Read your reply #19 again. Looks like you are suggesting an oval mono coil. I get better S/N with a figure8 Rx, reason I've posted figure8 shapes. Maybe I'm missing a figure8 shape I should be looking at.

[green]

Ordered some AWG30 magnet wire to see if better than AWG28. Some charts with AWG28. Targets chart straight line decay linear log after 3.5us. What might cause the decay signal to be lower between 2.5 and 3.5 us? Wondering if I should wind the AWG30 coils same number of turns.

[Carl-NC]

When you take data do you subtract a no-target response from the target response?

[green]

When you take data do you subtract a no-target response from the target response?

When I chart target decay I subtract the no target recording from the target recording.

Added a picture of part of the recordings.

[Carl-NC]

I thought you were, so if you're certain the data sets are correct then I don't know what might cause that. Usually you see an initial higher slope that dies out in about 1-tau, that's caused by the settling of the eddy currents to the perimeter of the target.

[green]

I thought you were, so if you're certain the data sets are correct then I don't know what might cause that. Usually you see an initial higher slope that dies out in about 1-tau, that's caused by the settling of the eddy currents to the perimeter of the target.

Have seen initial higher slopes in the beginning, especially with thicker targets. Was thinking the foil and AWG34 wire ring should chart straight line, foil (.016mm thick) AWG34 copper wire (.16mm diameter, .077us tau, 2inch diameter ring about 1.4us tau). Did another test this morning. Changed the initial balance. Forgot I had changed Rd from 1090 to 985 last night when trying to measure coil decay. The X10 probe across the coil lowers resonance. Data this morning was taken with Rd=985 instead of 1090 used for the prior test. Changed two things, Rd and initial balance (not smart). Need to do some more testing but first a question. During coil decay, the target signal is opposite polarity of eddy decay. Amplifier out signal goes negative when coil decay less than 2us then starts going positive. Wondering if amplifier out signal might be lower do to coil A/sec charging the coil, subtracting from eddy decay. Coil is around 3100A/sec at 2us and maybe 310A/sec at 2.5us. How low should Tx coil A/sec be to not subtract from eddy decay if it does? Any other suggestions what I might try?

Coil decay chart shows .17us tau, should be .217us

[green]

Looks like Rd isn't the cause. Maybe something with my coil.