I'm thinking about how I can make matching networks for the center of a
40 foot doublet for field day. I have a couple of those 20 footer
fiberglass poles and a center mount and want to be able to match on
20m, 15m, and 10m for now, and there are lots of ways I could do this,
ladderline to the base and a manual or auto tuner, autotuner at the
feedpoint, etc, etc, but what I'd really like to do is have switchable
fixed-tuned networks as my intent for this antenna is to be set up in
the middle of a field in the clear at the same height all the time, and
I want to have FAST bandswitching.

An autotuner really isn't in the cards financially, and needs power.

So, inductors are easy. I've got 500 feet of #10 copper wire.

I need capacitors. Preferably very very cheap ones, and I guess I need
a number of values.

So it seems like 3/4" copper tubing (type L) slipped into 5/8" copper
tubing with an air dielectric has a capacitance of 30pF per inch and a
breakdown voltage of 1300V, ballpark in the *ideal* case.

Two questions: It seems like having a small gap in a cylindrical
capacitor like this probably makes it act like more of a lumped-element
capacitor than, say, a length of RG-58. I guess the latter is more
like 30pF per FOOT, and so you need to use a much longer line in terms
of a wavelength to get the desired capacitance. Sound right to you
folks?

Second, I know the actual breakdown voltage of such an object is going
to be much less than 30kV/cm*the gap width. Anyone using copper tubing
caps like this (I saw a magloop "trombone" unit some years back like
this) and what are the dimensions and actual breakdown voltages of real
units?

I guess polishing the tubes, making the cut edges rounded and so forth
will help some, but maintaining the gap will be tough. I can easily
derate for concentricity errors, but what about the ends? What is a
good overlap of the two sections? Is it better to have the tubes be
exactly the same length, flush with each other, or maybe have one
longer than the other one so that the ends of one are far away from the
ends of the other, or is it likely no difference? I can always go to a
bigger gap; 0.035in diametral difference is probably pushing it for
100W use but I haven't yet calculated likely voltages across these
things, so here I'll take any general comments on making coaxial caps
that stand high voltage. I'd probably prefer air dielectric to teflon
heat-shrink or something, I don't know how available that is.

73,
Dan
N3OX

Right. So I figure if I don't need to ever exceed 10-15 inches of line
to build a capacitative stub on HF, it's better than if I need to get
out to many feet of line to get the same Xc.

So, here's a question. I've built a piece of such line since I posted.


The outer conductor is 15 inches long, 0.995 inch inner diameter. The
inner conductor sticks out on both ends, 1.5 inches on one end, 8
inches on the other (I've not cut it yet). The outer diameter of the
inner conductor is 0.875 inches. The dielectric is air, except for 1/2
inch at either end of the outer conductor where there are some little
nylon rings to keep the inner conductor spaced.

Incidentally, Zo of this line should be 7.7 ohms or thereabouts. I
expect the velocity factor is not so far from 1.

I attached it to an MFJ-259 via a small length (about 8 inches) of
RG-58 coax cable with a connector on one end and stripped ends (only
about an inch) on the other.

At low frequency (~2MHz) the analyzer shows a purely capacitative load,
about 200pF. (some of which is from the RG-58, some from the pipe)

However, this object (pipes+coax) has a series resonance around 30MHz!
(R=0,X=0).

This was not expected as the line is only 0.04 wavelengths long at this
frequency.

If I calculate Xc from Xc=Zo/tan(length in degrees) at 2MHz
(Zo=7.7ohm), I get Xc=482 ohms. C=165pF, not too far from measured,
considering that the 8 inch RG-58 probably adds a few tens of pF.

If I calculate at 30MHz where I'm finding a resonance, I expect Xc=31
ohms and Xl (Zo*tan(length in degrees)) = 1.88 ohms, so it should still
be strongly capacitative.

Where's my resonance coming from?

73,
Dan
N3OX

Right. So I figure if I don't need to ever exceed 10-15 inches of line
to build a capacitative stub on HF, it's better than if I need to get
out to many feet of line to get the same Xc.

So, here's a question. I've built a piece of such line since I posted.


The outer conductor is 15 inches long, 0.995 inch inner diameter. The
inner conductor sticks out on both ends, 1.5 inches on one end, 8
inches on the other (I've not cut it yet). The outer diameter of the
inner conductor is 0.875 inches. The dielectric is air, except for 1/2
inch at either end of the outer conductor where there are some little
nylon rings to keep the inner conductor spaced.

Incidentally, Zo of this line should be 7.7 ohms or thereabouts. I
expect the velocity factor is not so far from 1.

I attached it to an MFJ-259 via a small length (about 8 inches) of
RG-58 coax cable with a connector on one end and stripped ends (only
about an inch) on the other.

At low frequency (~2MHz) the analyzer shows a purely capacitative load,
about 200pF. (some of which is from the RG-58, some from the pipe)

However, this object (pipes+coax) has a series resonance around 30MHz!
(R=0,X=0).

This was not expected as the line is only 0.04 wavelengths long at this
frequency.

If I calculate Xc from Xc=Zo/tan(length in degrees) at 2MHz
(Zo=7.7ohm), I get Xc=482 ohms. C=165pF, not too far from measured,
considering that the 8 inch RG-58 probably adds a few tens of pF.

If I calculate at 30MHz where I'm finding a resonance, I expect Xc=31
ohms and Xl (Zo*tan(length in degrees)) = 1.88 ohms, so it should still
be strongly capacitative.

Where's my resonance coming from?

73,
Dan
N3OX

I guess I'm learning a public lesson but I'm going to post this as a
wrapup.

I figured out the problem. My inch of wire appears to measure about
0.06 microhenries.

1/(2*Pi*sqrt(0.06microhenries*165picofarads) ~ 50MHz

73,
Dan
N3OX

wrote in message
oups.com...
I'm thinking about how I can make matching networks for the center of a
40 foot doublet for field day. I have a couple of those 20 footer
fiberglass poles and a center mount and want to be able to match on
20m, 15m, and 10m for now, and there are lots of ways I could do this,
ladderline to the base and a manual or auto tuner, autotuner at the
feedpoint, etc, etc, but what I'd really like to do is have switchable
fixed-tuned networks as my intent for this antenna is to be set up in
the middle of a field in the clear at the same height all the time, and
I want to have FAST bandswitching.

An autotuner really isn't in the cards financially, and needs power.

So, inductors are easy. I've got 500 feet of #10 copper wire.

I need capacitors. Preferably very very cheap ones, and I guess I need
a number of values.

So it seems like 3/4" copper tubing (type L) slipped into 5/8" copper
tubing with an air dielectric has a capacitance of 30pF per inch and a
breakdown voltage of 1300V, ballpark in the *ideal* case.

Two questions: It seems like having a small gap in a cylindrical
capacitor like this probably makes it act like more of a lumped-element
capacitor than, say, a length of RG-58. I guess the latter is more
like 30pF per FOOT, and so you need to use a much longer line in terms
of a wavelength to get the desired capacitance. Sound right to you
folks?

Second, I know the actual breakdown voltage of such an object is going
to be much less than 30kV/cm*the gap width. Anyone using copper tubing
caps like this (I saw a magloop "trombone" unit some years back like
this) and what are the dimensions and actual breakdown voltages of real
units?

I guess polishing the tubes, making the cut edges rounded and so forth
will help some, but maintaining the gap will be tough. I can easily
derate for concentricity errors, but what about the ends? What is a
good overlap of the two sections? Is it better to have the tubes be
exactly the same length, flush with each other, or maybe have one
longer than the other one so that the ends of one are far away from the
ends of the other, or is it likely no difference? I can always go to a
bigger gap; 0.035in diametral difference is probably pushing it for
100W use but I haven't yet calculated likely voltages across these
things, so here I'll take any general comments on making coaxial caps
that stand high voltage. I'd probably prefer air dielectric to teflon
heat-shrink or something, I don't know how available that is.

73,
Dan
N3OX


Seems like you want something that once it is adjusted to the right
capacitance it will probably never have to be tweaked again. I would go with
brass plates. You can make up a nice little "brick" and then tweak it by
grinding parts of it off or drilling holes in it. Sounds a little crude but
really worked quite nicely for an antenna I built a long time ago. It did
prove a little impractical though, Capacitance really seemed to change a bit
with humidity. I never really solved this problem or even tried because I
came by one of the old Heathkit antenna tuners that was a motor driven cap
that you put in seies with a long wire antenna. Placing the brick in plastic
box would have probably been a viable fix. Also consider that your linear
caps may have quite a bit of inductance, just something else to figure into
it all.

Jimmie, I like the idea of drilling some holes to tweak the
capacitance. Good idea for a set-and-forget.... nice.

I've also used PC board caps. I don't know what the Q of those is
like; I guess it depends heavily on the dielectric... I'm sure the
duroid microwave board is much better than regular G10 or whatever.
Never really used them with power though.

There was a good tip in QST last? month about using big teflon heat
shrink tubing to make caps that were inherent parts of vertical
antennas, but I checked some prices of teflon heat shrink in large
sizes and it's not cheap.

73,
Dan

On 2006-06-16, wrote:
fixed-tuned networks as my intent for this antenna is to be set up in
the middle of a field in the clear at the same height all the time, and
I want to have FAST bandswitching.

An autotuner really isn't in the cards financially, and needs power.

So, inductors are easy. I've got 500 feet of #10 copper wire.

I need capacitors. Preferably very very cheap ones, and I guess I need


You have 500' of wire -- why bother with any switching/matching networks?

Just cut dipoles for each band and feed them off of the same feedline.


--
Alex/AB2RC

The short answer is because I like building antennas.

The more detailed answer is that I've got a forty foot long rotatable
doublet made from some of those 20' crappie poles, and I want to match
that. Making a fan dipole out of it would not really be possible. I'm
already going to do that for 40/80 for FD, but the doublet will work
well for the higher bands.

I'm eventually looking to have a 40-10m inclusive extremely lightweight
rotatable antenna with good efficiency on an easily erected mast. On
40m/30m I already have a loading and matching coil to make it a "shorty
40" style shortened dipole. For 20 and up I don't need loading I just
need to match it. It's going to be nice because it is tree
independent. I could set it up in a parking lot on a light mast if I
needed to.

Simple is not my style for field day. In a pinch I'll throw up simple
dipoles. In a real emergency, I'm not going to go fancy, I'm going to
go simple and resonant.

For field day, though, I like to put weird things up. I built a 3
element 15m yagi one year just for FD. The year before, I built an
aluminum tubing trap dipole for 10/15m.

If 15 and 10 are dead as I expect they might be, I might even turn the
crappie pole antenna into a delta loop for 20 only.

73,
Dan
N3OX

wrote:


If 15 and 10 are dead as I expect they might be, I might even turn the
crappie pole antenna into a delta loop for 20 only.

73,
Dan
N3OX


A fairly complete how-to for this is at

http://www.fros.com/KI0GU/w6zodelta.htm

tom
K0TAR

K7ITM wrote:
wrote:
I guess I'm learning a public lesson but I'm going to post this as a
wrapup.

I figured out the problem. My inch of wire appears to measure about
0.06 microhenries.

1/(2*Pi*sqrt(0.06microhenries*165picofarads) ~ 50MHz

73,
Dan
N3OX

Yeah, you discovered why they* suggest using an array of values of
bypass capacitors, even when using SMT capacitors with "zero" lead
length, to do a good job of bypassing a supply over a wide frequency
range. And why even in ham construction literature from the 1950s they
talk about keeping leads very short for good RF construction.
. . .

And just one of the many reasons I keep saying that good RF measurements
are much more difficult to make than most people realize.

Roy Lewallen, W7EL