Home

TX to Ae

Magnetics

L calc

Solenoid Z and Q

Paul Zwicky's solenoid Q optimisation
article in Funkamateur

Test coils with HP4342A Q-meter
Test coils with HP4342A Q-meter    -    Click picture to enlarge


An investigation of the solenoid inductance and impedance calculation techniques described in this website and elsewhere was given as an article by Paul Zwicky, HB9DFZ, in the German language magazine Funkamateur (Oktober 2013, p1080-1084, + picture on p1032):
"Optimierung der Güte einlagiger zylindrischer Luftspulen"
[ Optimising the Q of single-layer solenoid air coils ].

Paul has very kindly made his work files available as pdfs, with comments translated into English, and has allowed them to be published here. They are as follows:

Coil calculation with Medhurst and G3YNH .
Checks to see whether the coil calculation method holds against measurement.

Search for best Q while holding volume constant .
Scans through the main parameters to find the optimum number of turns while keeping volume constant

Algorithm for optimum coil and minimum volume .
Procedure to calculate mechanical data from an electrical specification.

Self-capacitance corrections .
The effect of C0 on Q factor.


In e-mail correspondence, Paul added the following:
"
I have also tried to use the method given in Arnold's paper ['The Resistance of Round-Wire Single-Layer Inductance Coils'. A H M Arnold. Proc. IEE Part IV. Institution Monographs, Oct. 1951. Vol. 98, p94-100]. In 2011 I converted all the necessary tables into approximations and was able to calculate ΦArnold. Then I did the same thing as in the file "searching for optimum". The result for ηopt(n) looked sensible. But αopt(n) looked too flat especially at small numbers of n. It even tended to have a maximum at >2, and that couldn't be true. Simply minimizing the wire length gives an αopt of 0.45. Looking into the Arnold paper we see that the verification is only by comparing with Medhurst. That means for n = 25 to 50. So I gave up.

Then I worked with Glenn Smith ['The proximity effect in systems of parallel conductors and electrically small multiturn loop antennas'. Glenn Smith. Dec 1971. Tech. Report No. 624. Division of Engineering and Applied Physics. Harvard Univ. Cambridge, MA 02138], which looked like very serious work for n = 2 to 8. I tried to extrapolate his findings for n > 8. I did not succeed well.

So I worked with your idea of finding Φ(n). And that was heureka!

Concerning capacitance:
My calculations ignored the presence of any capacitance. In the paper [Funkamateur, Okt. 2013] Dr Hegewald limits the validity to 1/5 of the SRF. It would be nice to have it more universal. On the other hand I carefully measured the C0 of the sample coils and used the findings to correct the direct measurement values L and Q. That means the values in my file are true for coils having no C0.

In order to give the limit of ≤1/5 of the SRF some meaning, I give the simple formula of Medhurst for C0 and calculate the SRF. Then the C of the connecting wires has to be added. The measurement of C0 and the calculation agreed to approx. ±5% ±0.2 pF.

To avoid additional losses I did not use any coil formers, and the coils were standing in free air on top of the measuring gear. Distance >D. [See p 1032 Funkamateur 10 2013, also see photo above].
"


HP4342A Q-meter with test coil
HP4342A Q-meter in operation    -    Click picture to enlarge



Home

TX to Ae

Magnetics

L calc

Solenoid Z and Q