Coil winding conundrum

[Andre' Kesteloot]

Talbot Andrew wrote:

> Moving from LF to MF for a bit, here's a problem for the topband
> operators and loading coil winders.
>
> At the weekend I made a loading coil to resonate the LF Tee antenna on
> topband.  34uH was needed and the only suitable former to hand was 40mm
> plastic drain pipe - the white type with no significant RF loss.
>
> First off I used 27 turns close wound of 1mm enamelled wire (which was
> all that I had of that wire), and completed the coil with 20 turns of
> 0.7mm diameter tinned copper with turns wide spaced at 3mm to allow
> tapping points.
>
> This worked fine but I noticed the coil got a bit warm with 100 Watts
> and decided I needed a higher Q....
>
> So, I made some Litz wire by twisting 20 strands of 0.25 mm diameter
> enamelled, which gives a copper X-sectional area a bit more than 1mm
> diameter equivalent.  Overall Litz wire diameter about 1.8mm.
> 40 turns of this, close wound with a few taps gave near enough the same
> inductance value for a winding length of 72mm
>
> Q was measured by connecting a 680pF cap in parallel, exciting the
> circuit with a one turn coupling coil driven from a synth, very loosely
> coupled and monitoring voltage across the coil with a x10 scope probe.
> Peak response frequency was noted and then the points either side where
> the response fell to -3dB (0.707) to get the bandwidth.  Then  Qu = CF /
> BW
>
> Now for the interesting bit.   When it came to measure the unloaded Q of
> each coil, the original one was a fair bit higher  at Qu = 140, compared
> with the Litz wound coil with Qu = 95.   Both coils were the same
> diameter, same inductance, and roughly the same length in total.   So
> why was the one made of plain wire better ?   Self capacitance ?    Q
> was only measured at 1 MHz, perhaps I should try measuring gain at a
> lower frequency.
>
> Puzzled and curious..................
>
> One interesting aside to come out though.  Rayners formula for
> inductance predicted the values of each coil to better than 10% and
> bears out observations I've made over the years  -  such a simple
> formula for wound inductors is too good to be true, working over a range
> in excess of  1000000 : 1
>
>         L(uh)  =    (D.N) ^ 2 / (460.D  +  1020 * length)     D = mean
> diameter, mm     N = turns
>
> Andy  G4JNT
>
> --
> The Information contained in this E-Mail and any subsequent correspondence
> is private and is intended solely for the intended recipient(s).
> For those other than the recipient any disclosure, copying, distribution,
> or any action taken or omitted to be taken in reliance on such information is
> prohibited and may be unlawful.