wrote:
If you change very much with a G5RV, it's not a G5RV
anymore. IE: If you feed a 102 ft dipole with ladder
line, but no choke or coax, it's not a G5RV anymore.

But that's exactly what G5RV recommended as one form of
his G5RV antenna.

Tell me this...What is the advantage of using the choke,
coax, etc, vs just running straight ladder line the whole
way?

The advantage is a pretty good match on 80m, 40m, 20m and
12m that's not guaranteed with straight ladder-line. Some
lengths of ladder-line present additional problems. For
instance, if one happens upon a current maximum point
located at a 4:1 balun, one can take 16 ohms down to 4
ohms. That's going in the wrong direction.
--
73, Cecil http://www.qsl.net/w5dxp

Owen Duffy wrote:
Is this the antenna described at http://www.qsl.net/w5dxp/G5RV.HTM ?

Nope, that's just an off-the-shelf vanilla G5RV.

In that article, ...

Forget that article which only shows why the *standard* G5RV is
a fairly well matched antenna on 80m and 40m.

Presumably when you say that the capacitor improves the VSWR on 75m,
you mean the VSWR on the coax. Did I miss something, how does the
capacitor improve the VSWR on 75m?

I'm sure you know this already. Given an SWR circle on a Smith Chart
that crosses the horizontal resistive line at less than 50 ohms and
given the 1/50 conductance circle, those two circles will cross at
two points. Where they cross in the capacitive reactance region is
the point on the transmission line where a parallel capacitance will
bring the impedance at that point to 50+j0 ohms. This is a common
matching technique for 75m mobile antennas. The same thing can be
done with a coil installed where the circles cross in the inductive
reactance region. This technique is described in the ARRL Antenna
Book.

What I have done on my G5RV is find the point where the SWR circle
intersects the 1/50 conductance circle in the capacitive reactance
region on 3.8 MHz and install a 1000 pf parallel cap there. My series
section line is 22.5 ft. of Wireman #554 at that point. The 50 ohm
SWR is reduced from about 5:1 to 1.3:1 on 3.8 MHz.

Given an SWR circle crossing the 1/50 conductance circle, there's a
point where a cap will result in 50 ohms. A little farther, a cap
will result in 300 ohms. A little farther, a cap will result in
450 ohms, etc. These are the points just past the current maximum
point where one can hang a capacitive stub to achieve a purely
resistive impedance.
--
73, Cecil http://www.qsl.net/w5dxp

Owen Duffy wrote:

On Tue, 10 Jan 2006 21:14:40 GMT, Owen Duffy wrote:


On Mon, 09 Jan 2006 15:50:20 GMT, Cecil Moore wrote:


Ricke wrote:

If SETUP right , There the best HF I've ever used.

Maybe the only one? :-) The G5RV, with tuner, is a pretty
good 80m, 40m, 20m, and 12m antenna. If the series section
is varied from 20 feet to 36 feet, it becomes a very good
all-HF-band antenna. With the addition of a parallel 1000pf
capacitor with the series section at 22 feet, on 75m my "G5RV"
has SWR of 1.3:1 and works as well as a 75m 1/2WL dipole.

Is this the antenna described at http://www.qsl.net/w5dxp/G5RV.HTM ?


I have made a mistake during my analysis, let me try again:

Now you tell me after I spent 15 minutes replying to it. :-)
The details are there so I won't repeat it here.

In that article, on 75m you model a feedpoint impedance of 36-j324,
28' of 300 ohm ladder line, for a Z of 15+j4 (seems to indicate 48.2
deg length of 300 ohm line with 0.007dB loss (optimistic)).

At that point, were 50 ohm coax connected directly, the VSWR at the
load end of the 50 ohm coax would be 3, however you shunt the 17+j4
with 1000pF to give a new Z of 17.3-j3.0 that results in a VSWR at the
load end of the 50 ohm coax of around 2.9, almost identical to the
case without the capacitor.

Presumably when you say that the capacitor improves the VSWR on 75m,
you mean the VSWR on the coax. Did I miss something, how does the
capacitor improve the VSWR on 75m?

What you missed is that the frequency must be changed to obtain the
benefit. The capacitor is *not* installed at the 17+j4 point. It is
installed at the 1/50 + j1/X admittance point. You can either increase
the length of the feedline past the 17+j4 point to the 1/50 + j1/X
admittance point or increase the frequency thus electrically lengthening
the feedline to the 1/50 + j1/X admittance point. You cannot keep both of
those values constant as you tried to do above.

You already know what I am trying to say. I must not be saying it
very well. When a parallel cap is used on a 75m screwdriver antenna
to achieve 50 ohms, the screwdriver is tuned to 1/50 + j1/X, i.e.
slightly inductive. When a parallel coil is used, the screwdriver
is tuned to 1/50 - j1/X, i.e. slightly capacitive.
--
73, Cecil http://www.qsl.net/w5dxp

On Wed, 11 Jan 2006 00:13:14 GMT, Cecil Moore wrote:

Owen Duffy wrote:
Is this the antenna described at http://www.qsl.net/w5dxp/G5RV.HTM ?

Nope, that's just an off-the-shelf vanilla G5RV.

In that article, ...

Forget that article which only shows why the *standard* G5RV is
a fairly well matched antenna on 80m and 40m.

But... in that article which recommends 28' of 300 ohm ladder line you
say "To improve the 75m SWR, try installing a 1000pF capacitor (mica
or doorknob) in parallel across the ladder line at the ladder line to
coax junction. Remove the capacitor for all other bands."

IMHO, just considering in isolation what is shown on that page there
is something inconsistent about the Smith chart, the impedances,
lengths, and assertions about the SWR improvement.

....

What I have done on my G5RV is find the point where the SWR circle
intersects the 1/50 conductance circle in the capacitive reactance
region on 3.8 MHz and install a 1000 pf parallel cap there. My series
section line is 22.5 ft. of Wireman #554 at that point. The 50 ohm
SWR is reduced from about 5:1 to 1.3:1 on 3.8 MHz.

This implies you are trying to "tune out" the shunt capacitive
reactance at a point on the line where the shunt resistive component
is 50... but you need the opposite sign of reactance reactance (so
that the susceptances subtract), you need an inductive reactance in
that case.

If "your G5RV" has a feedpoint impedance of 36-j324 (that seems
reasonable), your 22.5 ft. of Wireman #554 will transform that to
21.53-j53.33, and the VSWR in 50 ohm line connected at that point
would be 5.3.

A shunt capacitance CANNOT improve the 50 ohm VSWR at that point

The effect of the shunt 1000pF capacitance is to change the impedance
at the junction to around 3.6-j25, which would cause a VSWR of around
17 in the 50 ohm line.

However:

If the ladder line was around 31' in length, then the Z at that point
would be around 21+j25 (equivalent to 50 ohms R in parallel with +43
ohms X), and a shunt 1000pF (~ -42 ohms X) capacitor would give nearly
perfect VSWR on the 50 ohm line.

In summary, in a general sense, if you want to use a shunt capacitor
as you propose, you need to find length of line such that the
admittance at that point is 1/50-jB (negative susceptance is
inductive), and the correct shunt capacitor has a reactance of 1/B.

Flawed explanation aside, the only way that 22.5' works is if your
feedpoint Z is quite different to 36-j324.

Owen
--

On Wed, 11 Jan 2006 00:36:25 GMT, Cecil Moore wrote:


You already know what I am trying to say. I must not be saying it
very well. When a parallel cap is used on a 75m screwdriver antenna
to achieve 50 ohms, the screwdriver is tuned to 1/50 + j1/X, i.e.
slightly inductive. When a parallel coil is used, the screwdriver
is tuned to 1/50 - j1/X, i.e. slightly capacitive.

Negative susceptances are inductive. An inductive reactance of j5 is a
susceptance of 1/j5 or -j1/5.

I agree with your words, the sign of the admittances is wrong.

Owen
--

Owen Duffy wrote:
But... in that article which recommends 28' of 300 ohm ladder line you
say "To improve the 75m SWR, try installing a 1000pF capacitor (mica
or doorknob) in parallel across the ladder line at the ladder line to
coax junction. Remove the capacitor for all other bands."

Yes, but installing the cap will raise the resonant frequency.

What I have done on my G5RV is find the point where the SWR circle
intersects the 1/50 conductance circle in the capacitive reactance
^^^^^^^^^^
region on 3.8 MHz and install a 1000 pf parallel cap there. My series
section line is 22.5 ft. of Wireman #554 at that point. The 50 ohm
SWR is reduced from about 5:1 to 1.3:1 on 3.8 MHz.

This implies you are trying to "tune out" the shunt capacitive
reactance at a point on the line where the shunt resistive component
is 50... but you need the opposite sign of reactance reactance (so
that the susceptances subtract), you need an inductive reactance in
that case.

Sorry, I misspoke. Where I said "capacitive reactance region"
above, it should have been "inductive reactance region".

If "your G5RV" has a feedpoint impedance of 36-j324 (that seems
reasonable), your 22.5 ft. of Wireman #554 will transform that to
21.53-j53.33, and the VSWR in 50 ohm line connected at that point
would be 5.3.

"If" is the important word. My G5RV is obviously different from your
values. It's made out of insulated wire and I'm not sure it is exactly
102 feet long.

Flawed explanation aside, the only way that 22.5' works is if your
feedpoint Z is quite different to 36-j324.

And it is obvious that's the case. The "450" ohm ladder-line is 22.5 ft.
long and a 1000 pf capacitor resonates it on 3.8 MHz. Whatever the
feedpoint impedance needs to be to cause those conditions, that's
what it is.
--
73, Cecil http://www.qsl.net/w5dxp

Owen Duffy wrote:

Cecil Moore wrote:
You already know what I am trying to say. I must not be saying it
very well. When a parallel cap is used on a 75m screwdriver antenna
to achieve 50 ohms, the screwdriver is tuned to 1/50 + j1/X, i.e.
^ should be -
slightly inductive. When a parallel coil is used, the screwdriver
is tuned to 1/50 - j1/X, i.e. slightly capacitive.
^ should be +
Negative susceptances are inductive. An inductive reactance of j5 is a
susceptance of 1/j5 or -j1/5.

I agree with your words, the sign of the admittances is wrong.

Yes, you are correct - sorry. But it now seems that you understand
what I was trying to say. If one takes an ordinary G5RV and
installs a parallel 1000pf capacitor at the coax/twinlead junction,
one will raise the resonant frequency and lower the SWR on the
coax for 75m operation. Very close to 50+j0 ohms can be achieved
on 75m through that simple act. When I lived in AZ, I switched that
cap in automatically using a relay and the frequency output signal
on my IC-745.
--
73, Cecil http://www.qsl.net/w5dxp

On Wed, 11 Jan 2006 04:25:59 GMT, Cecil Moore wrote:


Yes, you are correct - sorry. But it now seems that you understand
what I was trying to say. If one takes an ordinary G5RV and
installs a parallel 1000pf capacitor at the coax/twinlead junction,
one will raise the resonant frequency and lower the SWR on the
coax for 75m operation. Very close to 50+j0 ohms can be achieved
on 75m through that simple act. When I lived in AZ, I switched that
cap in automatically using a relay and the frequency output signal
on my IC-745.

OK.

I played around a bit using the feedpoint impedances that I modelled
for my "Feeding the G5RV" article. With 31' of 554, I needed about
2000pF to "tune" it for low 50 ohm VSWR at 3.6MHz.

I plotted the impedance presented to the coax for a range of
frequencies from 3.5 to 3.8MHz, they are at
http://www.vk1od.net/temp/G5RV-W5DXP.GIF . The Smith chart is
normalised to 50 ohms. The solution seems fairly narrow band, the VSWR
at 3.55 was 6, at 3.6 it was 1.3, and at 3.65 it was 5.

Of course, implementations will have slight differences in actual
feedpoint impedances, and the outcome is very sensitive to slight
differences in feedpoint Z. This "no-tuner" matching scheme will
probably need significant customisation for each implementation.

Owen
--

"Owen Duffy" wrote
Of course, implementations will have slight differences in actual
feedpoint impedances, and the outcome is very sensitive to slight
differences in feedpoint Z. This "no-tuner" matching scheme will
probably need significant customisation for each implementation.


==========================================
The World-famous G5RV.
---------------------------------
What everyone appears to forget, is that Zo of the balanced twin-line
section, on all bands except at 14.15 MHz, has a considerable affect
on feedpoint impedances, swr, losses, etc.

When describing systems and performance nobody ever mentions what Zo
of the feedline actually is. Omission of Zo reduces any following
discussion to blythe, innocent nonsense.

R.L.Varney himself never gave a value to Zo. He didn't need to. He was
concerned mainly with 14.15 MHz. It would be unfair to accuse him of
not understanding the serious effects of Zo on other bands.
----
Reg.

Owen Duffy wrote:
I played around a bit using the feedpoint impedances that I modelled
for my "Feeding the G5RV" article. With 31' of 554, I needed about
2000pF to "tune" it for low 50 ohm VSWR at 3.6MHz.

I'm just reporting what it took for my actual antenna under the
existing conditions at my QTH. The cap is actually 950 pf for
a minimum SWR of 1.3:1 on 3.9 MHz. The optimum value of the cap
would no doubt change at lower frequencies. With 22.5' of
Wireman #554 and a 950 pf cap, the 3:1 bandwidth is 145 kHz.
Adding sections of ladder-line lowers the resonant frequency.

Incidentally, this is a method for modifying the G5RV to work,
not only without a tuner, but with built-in tuners. When using
a built-in tuner, the antenna configuration doesn't have to
be changed as often. My IC756PRO will tune my present configuration
from 3.72-4.0 MHz. or 280 kHz.

I plotted the impedance presented to the coax for a range of
frequencies from 3.5 to 3.8MHz, they are at
http://www.vk1od.net/temp/G5RV-W5DXP.GIF . The Smith chart is
normalised to 50 ohms. The solution seems fairly narrow band, the VSWR
at 3.55 was 6, at 3.6 it was 1.3, and at 3.65 it was 5.

Changing the length of the series section will shift the resonant
frequency. I can vary mine from 22.5 ft. to 38.5 ft for a near-
perfect SWR on all HF ham frequencies.

Of course, implementations will have slight differences in actual
feedpoint impedances, and the outcome is very sensitive to slight
differences in feedpoint Z. This "no-tuner" matching scheme will
probably need significant customisation for each implementation.

IMO, that is what ham radio is all about - warm up the old MFJ-259B
and get with the program. :-)
--
73, Cecil http://www.qsl.net/w5dxp