I just tried a quick model to see if a single loading coil at the center
of a 40m dipole element was feasible. For simplicity I modeled a 46-foot
element with constant 0.75" diameter. I compared two loading coils
located halfway out each half-element with one loading coil at the
center. I was expecting the single-coil element to have much narrower
SWR bandwidth. But SWR for the two designs was remarkably similar. I
adjusted the inductances for resonance at about 7.15 MHz. The SWR of
two-coil design was about 2.7 at 7.0 and 7.3 MHz when matched at 7.15
MHz while that of the single-coil design was 2.85. Both had a load loss
of 0.06 dB using coil construction similar to the M2 coil (calculated Q
about 965). The inductance of the single coil was 8.0 uH while that of
the two coils was 7.7 uH each. I quote SWR values only to indicate the
inherent Q of the elements. They are not what a Yagi made from such
elements would exhibit.
In addition to simplicity, the advantage of using a single coil is that
when optimized for maximum Q (about 1230), the diameter increases to
about 6.9" with a length of about 4.7". Enclosing a coil of this size to
keep the Q from degrading when wet would create two large wind loads
halfway out each half-element. But the wind load would be no problem
when mounted at the boom. VE6WZ seems to get by without coil enclosures,
but these results are alarming:
http://www.n3ox.net/tech/coilQ/
After examining the M2 coil manual and making measurements on the coil
illustration, I estimate the coil diameter to be 3". The length is
5.8125" according to the description (15.5 turns of 3/16" tubing spaced
the wire diameter). Lead length is 1.5" to the element centerline. All
dimensions are wire center to wire center. You can model the coil
inductance and automatically optimize Q with this:
http://ham-radio.com/k6sti/coil.zip
See README.TXT for installation instructions. I will be updating the
program tomorrow, but the current 3.89 version works fine.
I used the RLC coil model in my antenna model. Inductance varies
somewhat over 7.0 to 7.3 MHz due to coil self-resonance so there is some
error when you use a simple RL model to cover the whole band. The L and
C of the RLC model are constant over 40m. I saw some difference in SWR
at the band edges between the RL and RLC models though nothing great. If
your antenna modeling program can handle RLC loads (R in series with L,
C in parallel with that series combination), use it for best accuracy.
Incidentally, Copperweld makes copper-clad aluminum wire if you want to
reduce coil weight when mounting coils halfway out each half-element.
AWG 5 wire (0.1815") with 10% copper by weight has a copper thickness of
4.7 skin depths at 40m. That puts 99% of the current in the copper. The
M2 coil uses copper-clad aluminum.
https://www.copperweld.com/application/files/6815/3833/2604/Welded_Copper-covered_Aluminum_CCA_10.pdf
I'm hoping someone verifies this loading comparison. The results are
surprising and I'm always suspicious of unexpected favorable results.
Brian
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|