In article ,
David wrote:

Dave,

If I scale that up to 915MHz, the dimensions are almost perfect for
using RG58 coax instead of the copper tubing.The outer tube would be
5.5mm and inner conductor 1.6mm. Being 50 Ohm coax, I assume therefore I
do not need to slide the tap point up as their design did.

Bad assumption, I think. The portion of the matching section between
the tap point and the bottom is a shorted stub, which adds some
inductive reactance at the feedpoint. If you don't have it, you won't
get a good match.

Another approach I've seen to creating the necessary shorted stub is
to just use RG-58 and a simple BNC "T" connector. If you look at the
Sperrtopf antenna design, picture it made this way:

- one long piece of RG-58 1380 mm long, with the upper part of
the outer shield removed... that's your radiator, and the portion
of the matching section above the feedpoint.

- A second section of RG-58, about 100 mm long, with one end shorted.
That's the stub.

- The feedline RG-58.

Now, just connect all three together... radiator/matching section,
stub, and feedline. You can solder 'em, or add BNC plugs and use a
BNC "T". If you do the latter, remember to include the lengths of the
plugs and "T" into account.

I suppose I
should also add the choke coil at the antenna base for additional
isolation of ground currents.

Certainly would not hurt to do so. You could either coil the coax
below the bottom of the matching section, or add a quarter-wave choke
sleeve, or just run the coax through a few ferrite beads.

The article mentions the top part as 1/2 wave and bottom at 1/4 wave. If
the bottom part of their design is 505mm then the antenna must be used
for 2m band. If this is the case, how come the radiating element is 1480
long ?

I think there's some confusion in the antenna schematic diagram,
compared with the text and parts list. The diagram appears to show
the upper (radiating) portion of the center element being 1480 mm, but
the parts list indicates that the *total* length of the inner element
is 1480 mm.

The latter seems more correct to me... it'd give you a radiating
element of (1480 - 505) = 975 mm, which is a bit less than 1/2
physical wavelength at 2 meters. This seems reasonable to me given
that the radiating element is fairly thick (10mm) rather than a thin
wire.

The Coaxial antenna mentioned in the article mentioned by Jerry looks
good. The version they mention with a GAP would be simple to construct.
It is a 1/4 wave radiating element at the top (earth braid stripped
back), then 1/4 wave of full coax, then a "small" gap, then the
transmission line. They do not mention the size of the gap but I assume
any small gap has the same effect of isolating the reverse current ?

I'm not clear about the gap configuration either. The article sort of
implies that its size has to be set experimentally, so that the
leakage current through the gap is the right amount to cancel out the
signal propagating back down the outside of the upper section. The
article doesn't say whether the gap is supposed to go around the full
circumference of the coax (completely isolating the upper and lower
portions of the shield) or only partially around... I infer that it's
the latter, because I don't see how the antenna could work with the
shield sections fully separated.

Should the actual lengths of these sections be modified by "K" depending
on diameter of conductor or are they exact 1/4 wave length cuts ?

Well, I think it's a bit of a compromise.

The upper section (the coax center conductor and its surrounding
dielectric) is going to have a velocity factor of not much less than
1.0. Cutting to 1/4 physical wavelength, and then trimming a bit,
seems likely to work.

I think the lower section is likely to be a bit trickier, since
there are actually two velocity factors involved. The radiating is
done by the RF travelling back down the outside of the coax braid from
the feedpoint, and this (like the upper section) will have a velocity
factor close to 1.0, which suggests that the gap-to-feedpoint distance
should be close to 1/4 physical wavelength.

HOWEVER: the signal travelling up the *inside* of the braid, like that
travelling up the inner conductor towards the feedpoint, will be
travelling more slowly - the velocity factor will be somewhere around
..66 - .8 depending on cable type. This means that there will be more
than 180 degrees of phase delay between the signal passing the gap
going upwards (inside the cable) and the signal reaching the gap from
above (travelling back down the outside of the cable). Hence, the gap
leakage won't result in full cancellation of the signal travelling
down the outside of the coax past the gap.

To get the two signals into accurate 180-degree phase opposition,
you'd need to make the distance from the feedpoint to the gap a bit
less than 1/4 physical wavelength, but a bit more than 1/4 of the
coax's usual electrical wavelength. Splitting the difference might
work fairly well.

I'm not sure what this will do to the antenna's feedpoint impedance /
return loss, though, and it'll probably tilt the antenna's radiation
pattern somewhat. You will probably need to experiment to get the
best compromise between antenna pattern and feedpoint impedance /
return loss / SWR. The same is true of the gap size and configuration.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!