PCB Coils & MORE

[Sean_Goddard]

Not quite sure who understands how to design coils properly, but from what I've seen...NO ONE!!

I say this because I have just spent four hours, downloading around 20 coil calculators and no two gave anything like the same results when fed with the same data. Conclusion?

1) It's a black art and I haven't sacrificed enough virgins (I DO live in the UK though )

2) No one has the faintest how to implement the equation properly

3) Everyone who has written a calculator is cr@p at maths

That aside, I have a suggestion, which is;

Taking Dave Emerys excellent co-planar coil design article, can some whizz come up with a piece of software (maybe written in VB) which calculates the TOTAL inductance of a coil, based on the Tx coil AND the Bucking coil? That is to say a program which will calculate the whole caboodle.

Give the program the desired operating frequency, coil size (inner / outer), specify the type of oscillator used (Colpitts or Clapp etc) then generate the whole lot to enable you to simply wind the damned coil which would entail;

No of turns, recommended wire dia. former diameter, height of coil for OUTER Tx. SAME for RX coil and SAME for Bucking coil then also give the values of C for the oscillator.

The SAME could be done with a PCB based solution ( http://www.edn.com/design/components...iral-inductors ) you could then put the oscillator ON THE COIL PCB and make an X-Terra type machine. Change coil to get different frequencies.

THB I could do this, but I simply don't have the time. Perhaps Aziz or one of our other resident Genii could do something about this.

Even maybe OVAL Rx coils (Catseye) to give the best of DD and Concentric. Come on, someone give it a go!!

I think Aziz could even work out how to balance a PCB coil too .

[Qiaozhi]

I say this because I have just spent four hours, downloading around 20 coil calculators and no two gave anything like the same results when fed with the same data.

That's because they all use slightly different equations and/or make certain assumptions. The least accurate coil calculators are the ones based on Wheeler's equation, as this is designed for calculating the inductance of solenoids, and not air-cored coils for metal detectors. The most accurate results are obtained by using Brook's equation. Assumptions made here are that the coil has a square cross-section, but in practice the coil is bundle-wound with little consistency over the tightness of the bundle. The bottom line is that a good calculator will get you into the right ballpark, but you will still need to make some adjustments to achieve the desired result.

Also remember, when designing a concentric coil, that the bucking coil inductance needs to be subtracted from the TX inductance as it's in anti-phase.

[Silver Dollar]

I worked at a company that made an EM device. We had PCB based coils made (at great expense) and had problems
with shorts appearing and disappearing messing up the data. We had to bake them at 60C for a day or two then sort out
the good ones. At the end we had a pile of about $40000 worth of useless coils!

I imagine the real world has too many changing variables to get accurate calculations for inductance and capacitance.
I read somewhere that most caps value varies with temp and time and they have a lifespan that tries to keep them in
spec for x amount of time. My own measurements with a coil show lots of changes day to day.

I bought some wire off ebay and it was speced to be a certain AWG but measured a couple gauges smaller. All the coils
I made with that wire were off a bit.

I have a new problem as I wound my (many) coils before finding a suitable coil case. Now I have a case and the coils don't fit!
So I have to calculate how many windings to make the new size coils...

The best I've come up with is to wind a coil and measure it, then fudge the numbers until the calculator reads the same, now
you can change the number of windings to get different inductances in the same size...