Date: Mon, 23 Feb 2015 08:51:41 -0800
From: Jim Lux <jimlux@earthlink.net>
To: towertalk@contesting.com
Subject: Re: [TowerTalk] Mosley Antenna Question
On 2/23/15 7:24 AM, Jim Thomson wrote:
## Just thought of something else. Putting a coil inside of an aluminum
cylinder will decrease the inductance of any coil.
I'm not sure it actually changes the inductance.. The parasitic C in
parallel can change the "end to end" reactance so that it looks like L
is lower.
And, since the can forms a "shorted turn" (assuming it's not slit down
the length), and that turn is coupled to the coil, the apparent
inductance will change. probably lower, but it depends on the mutual
inductance of the can and the coil (which might be quite high: the
magnetic flux is fairly well coupled since the "turns" are concentric.
## So now you need more turns to get the same uh.
> Wind small gauge AL wire onto a grooved polystyrene form, encased by an AL
> tube,
> and you now have the worse possible config for a coil.
polystyrene isn't a great dielectric in terms of loss, but it won't
change the inductance. It will change the parasitic C (makes it
bigger), and to the extent that the E field of that C interacts with the
dielectric, there's increased loss.
Aluminum wire is an issue, but I think "small gauge" might be a more
important one than the material. After all, we don't copper or silver
plate our HF antennas.
Skin depth in copper at 30MHz is 12 microns, 15 microns for Al (25% bigger)
The resistivity of aluminum is 50% higher than copper, so combining the
increased skin depth and the increased skin depth, the net for Al vs Cu
is that Al has 20% more loss than Cu (at 30 MHz)
Skin depth goes as sqrt(resistivity) and sqrt(1/f).. going lower in
frequency doesn't change the overall calculation: the skin depth gets
bigger, but the ratio stays the same. At some point, when the wire get
very small (<5 skin depth in diameter), then it's not skin depth any more.
AWG30 wire (pretty fine) is 0.01" in diameter, 254 microns, so at HF
frequencies, pretty much any wire you use is going to be a "large"
conductor, and resistance (or loss) goes as the 1/diameter, not 1/area.
### I have spare airdux laying about, aprx 2 inch od, in aprx 12 ga and also
14 ga tinned.
Also have some airdux that is huge stuff, 4 inch OD, made from 8 ga copper, not
tinned.
Also have some hb .25 inch cu tubing coils laying about, 1.5 inch ID and 2.0
inch OD.
## took some AL pipe, 2.875 inch OD, sched 40.....and slid it over both the 2
inch OD airdux
coils and also the hb tubing coils. What a disaster, uh drops. Q drops.
If you intend to
use a coil for loading an ant...or in a linear amp, I would not suggest
sliding AL pipe or tubing
over top of it anytime soon.
## On a trapped triband yagi, the 10m trap is used as a loading coil on 15 +
20m. The 15m trap
is used as a loading coil on 20m.
## In trap mode, it’s a parallel resonant circuit, with the C being the C
between the AL tube slid
over the coil..and the coil itself. ( telrex used a separate outboard HV cap,
npo style, like a HT-50 /58 /57. )
## IMO, the trap would work better if an outboard cap was used. The
current through the coil is just wicked,
when in trap mode. Peak V across the...C between al tube and coil is also
wicked.
## Unloaded Q of a coil is just its XL / ESR. For a fixed freq, say 7 mhz,
if we double the uh of the coil, by
making it aprx twice as long, both the esr and the uh will double. Q wont
change.
## skin depth goes to the sq rt of the freq. .25” tubing coil will handle 41%
more current on 80m vs 40m.
(sq rt of 2 = 1.414).
.25 inch tubing will handle 4 X more current on 40m...vs 10m. ( sq rt of 4 =
2).
## the kicker is of course.... XL goes up in proportion to freq. Our 40m
tubing coil will have the same uh on 10m
as it does on 40m. But XL=2pi X FL. So XL on 10m is quadruple what it is on
40m. Meanwhile the skin depth on
10m is only one quarter of the same coil..on 40m. Q of the coil on 10m is
now DOUBLE vs the same coil on 40m.
( XL went up 4 X.... but esr only doubled)
## So now we have this nice 10M coil with its sky high unloaded Q. Sounds
great..but the fact is, its ESR is 4 x
higher. The high esr is what kills it. The 10m coil material needs to be 4
X bigger than the 40m coil material.
This is why 10m tank coils are bigger material. On a side note, strap
conducts RF on both sides...where as tubing
only conducts on the outer surface. In some applications where u need small
values of UH, and also high current
handling capacity, coils wound from strap, like a 10m tank coil, will work a
lot better, and very easy to wind.
1 inch wide strap has a total circumference of 2 inchs = .637 inch OD tubing.
To make the small .4 uh coils
in my 40m ssb /cw relay box, I used a few turns of 1” wide silver plated cu
strap, .032 inch thick, wound on a
1.5 inch ID. Beauty is, 2 of them fit nicely inside a nema box. C between
turns is zero. Its like two knife edges
facing each other. Closest tubing to .637 inch is .625 inch od. Winding
.625 tubing coils is no fun, plus its total OD
is huge vs the 1.5 inch ID strap coil.
## ESR of Al is 1.645 greater than Cu. That’s a huge difference. If Al is
used, it has to be 1.645 greater OD material
vs Cu. That info came from k6sti coil.exe software. That was in the case
of 2.8 uh coils for loading a 340N.
(Im experimenting with replacing the oem .187 inch solid al LL rods with .375
inch coils). 4.5 turns of .375 inch
AL vs .375 inch Cu. ESR is .125 ohm vs .076 ohm. (.070 ohm if silver
plated) Q is 1001 / 1653 / 1791 respectively.
## Mosley rates its 33jr coils for 250w. it rates all its bigger tribanders
for only 600w rtty. The current has to be sky high
to cook the Al wire used in trap coils. In trap mode, there is a lot of
circulating current through the coils. On 20m,
both the 10+15 m traps are in loading coil mode. The ones I saw melted, were
from extensive use on 20m, with too
much applied power. Traps will work, if built correctly. Cu wire...( or
bigger AL wire) + external npo caps might do the trick..albeit the
npo caps would be expensive.
Jim VE7RF
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|