Coil Calculator

[torrente1976]

I hope you help build your dishes greetings to all the brothers aficcionados decteccion

http://coil32.net/download.html

[Qiaozhi]

I hope you help build your dishes greetings to all the brothers aficcionados decteccion

http://coil32.net/download.html

This is a solenoid coil calculator, whereas metal detectors (in general) use multi-layer air-cored coils. It's probably based on the usual Wheeler's equation.
We already have an air-cored coil calculator here -> Coil Calculator located as a "sticky" at the top of the Coils Forum, which is based on the more accurate Brooks equation.

[Skippy]

There appears to be a number of different coil formats (bobine) on the left-side menu, maybe the poster can translate (into English, of course) these choices?
If it uses Maxwell's equations, (as stated), then does this mean it does not take into account mutual inductance between adjacent wires in a bundle? Ie. a 10 turn small-bundle coil is just seen as 10 repetitions of a single turn?

[kt315]

for whom got the DVD-ROM with archive open a folder '#coil_calculator_soft' - you will see there a diggest of the calculators' soft.

[Qiaozhi]

There appears to be a number of different coil formats (bobine) on the left-side menu, maybe the poster can translate (into English, of course) these choices?
If it uses Maxwell's equations, (as stated), then does this mean it does not take into account mutual inductance between adjacent wires in a bundle? Ie. a 10 turn small-bundle coil is just seen as 10 repetitions of a single turn?

Actually you're correct. I've just found the English description, and it appears to be based on Maxwell's equations.

If you download the Linux version and look at the Python code, you can see how the calculations are done. Unfortunately I don't have the GTK libraries loaded, so I reverted to downloading the Windows executable.
Firstly I ran our Coil Calculator using a coil radius of 127mm, wire thickness of 0.65mm, and 20 turns of wire. The result was a mean radius of 128.45mm, a coil thickness of 2.91mm, and an inductance of 302uH. From experience I know that this result is very close to reality.
Using the Coil32 program, I entered the diameter of the former as 254mm, a winding length of 2.91mm, thickness of coil as 2.91mm, wire diameter of 0.65mm, and a wire diameter with insulation of 0.65mm. The result was an inductance value of 130.574uH to 216.08uH (number of turns of the coil: 17 to 20). This is clearly an incorrect result.
I can't see that I've entered anything incorrectly, but perhaps someone else can check as well.

[radioacoustick]

I'm developer of the Coil32 calculator and I would be glad to join to the discussion of my program. I think, it can help me to improve the algorithm of the app. Calculation of multilayer inductor is based on Maxwell's equations of mutual inductance between two coaxial circular conducting filaments. More about this algorithm:
http://coil32.net/multi-layer-coil.html
http://electronbunker.ca/eb/CalcMethods1c.html
http://www.coe.ufrj.br/~acmq/tesla/maxwell.pdf

[radioacoustick]

I can't see that I've entered anything incorrectly, but perhaps someone else can check as well.

Previous my post on moderation. I'm developer of Coil32 and I would be glad to join to discussion.
If we try to enter actual input data on "Coil" tab, we'll have the following result:

1 ----------------------------------------------------------------------------
Coil calculation
Multilayer coil
Coil inductance 302 µH
Diameter of former D: 254 mm
Winding length l: 2.91 mm
Wire diameter d: 0.65 mm
Wire diameter with insulation k: 0.65 mm
=>Thickness of the coil c: 3.9 mm
=>Resistance of the coil 0.896 Ohm
=>Necessary length of wire: 16.984 m
=>Number of layers: 6
=>Number of turns of the coil: 21

We have 6 layers, 5 layers with 4 turns in every one and 6-th layer with a single turn. In this case the coil thickness is bigger then double coil mean radius. Perhaps, the algorithm of result data averaging in "Inductance" tab have need for improving. This is the case of boundary going up of layer number.
In common case, I think this algorithm is better that any simple formulas (Wheeler's, Brooks etc.) because it have independence of winding geometry, another words - "short" or "long" coil we have.

[Qiaozhi]

Previous my post on moderation. I'm developer of Coil32 and I would be glad to join to discussion.

Hi radiocoustick. You are more than welcome here!
I have approved your initial post, which is now out of moderation. If you include attachments or links when you make your first post, it always goes into moderation awaiting approval. This is to prevent spammers having easy access to the forum.

[Qiaozhi]

We have 6 layers, 5 layers with 4 turns in every one and 6-th layer with a single turn. In this case the coil thickness is bigger then double coil mean radius. Perhaps, the algorithm of result data averaging in "Inductance" tab have need for improving. This is the case of boundary going up of layer number.
In common case, I think this algorithm is better that any simple formulas (Wheeler's, Brooks etc.) because it have independence of winding geometry, another words - "short" or "long" coil we have.

When I tried to compare the results of Coil32 with the Geotech Coil Calculator, I was attempting to produce a square cross-section to the coil, as this is the assumption with the Brook's equation. However, in reality, the exact geometry of the coil can deviate quite significantly before the accuracy of the calculated inductance becomes unacceptable.
No doubt the Coil32 program is capable of producing a more accurate result, but in practice there are many factors that result in a "real" coil being different from the theoretical values. For example, when a coil builder constructs the "real" coil, the inductance is very dependent on how tight the winding are bound together. Also, the inductance can vary somewhat once the windings are potted. At the end of the day, coil calculators are useful for getting you into the right ballpark, so to speak. Quite often there's still some tweaking required.

[radioacoustick]

Yes, I agree with you. Any formulas and algorithms are based on some idealistic model of winding. Even we try to take the program that base on fundamental Maxwell's equation of electromagnetic field, we don't take the surefire result that would be equal to reality. It seems, from comparison of the results, the Coil32 was getting right ballpark in "Coil" tab, but wasn't getting one in "inductance" tab. I think, this ballpark needs bigger area.

[homefire]

This is why The Great Creator had us make Inductance meters. You can Twist up the coil as folks tell you it should be and Clamp into the wire with little Damage and see ware your at. Clamp marks can be fixed with a bit of Nail Polish.

[Dave J.]

WORLD'S SIMPLEST COIL CALCULATOR for regular "air core loop" coils
The coil can be round, D-shaped, square, elliptical, etc. Can't be a solenoid or a "pancake coil".

The magnetic constant in electromagnetic physics is 4pi x 10 exp-7 Henries per meter. Approximately 1.3 microHenries per meter.

Multiply the circumference of the coil in meters by 1.3 uH and by the square of the number of turns. That's your estimate of the inductance of the coil you'll be winding. So go ahead and wind it.

After you've wound the coil, measure the inductance. If it isn't the inductance you were aiming for, whatever the percentage difference is, cut that number in half and add or subtract that percentage number of turns. For example if the inductance was 18% off your goal, add or remove 9% of the turns. You'll be very close to the inductance value you wanted.

[Skippy]

My solution is to wind a few extra turns, then when the coil is in its final shape, and tightly bound up / glued etc, join in the extra turns as needed. Example wind (40T + 2T +1T ) allows you to adjust from 37T up to 43T, giving about +/-15% adjustment. This gets around problems such as when you wind a coil with loose bundle, and the final tightening increases the L value.

[Qiaozhi]

All the coil calculators tend to slightly underestimate the value of the coil inductance, which is why the method outlined by skippy is the same as I use. The World's Simplest Coil Calculator appears to overestimate by a fair margin, and it's a lot more difficult to remove windings than it is to add more. Leaving an extra winding or two is also a very useful method when coil balancing.

[Dave J.]

George is right. "The world's simplest coil calculator" does best on a coil that's got a tight bundle (small cross-section) relative to the shortest dimension across the coil. A thick or loose winding or a narrow coil shape that brings opposite sides close to each other will drop inductance by quite a bit compared to the calculated value.

When creating a new coil design, I usually just wind the mandrel with wire of any convenient gage to the expected thickness of the bundle, with a known number of turns. Then measure the inductance. If it's a nonmetallic mandrel the L-measurement can be done right there on the mandrel without even bonding the wire as long as the windings are not loose. Dividing the inductance by the square of the number of turns gives me the one-turn inductance factor for that mandrel, and from then on out it's easy to wind to whatever inductance you want and to know what gage wire and how many turns to do it with.