Topband: Inverted L Matching Summary

W4nf@aol.com W4nf@aol.com
Mon, 29 Jan 2001 22:38:07 EST


A lot of you asked for a summary of the responses.  Here is a summary of the 
responses I received.
_______________________
Dear Jack,

I use a 1000 pfd variable capacitor (A Jennings UCLS-1000) to ground, and a 
B&W
3" O.D.
coil stock in series. The series inductor is tapped for resonance at 1.830 
MHz,
and the capacitor to ground is tapped at an appropriate place on the inductor 
and
tuned against ground for lowest SWR.

My antenna is a inverted "U"; up 60 to 70 feet; out 100 feet; & down 60 to 70
feet to ground. There is a 100 foot wire from the feed end ground to the
termination ground. I have 12 radials at the feed end, and 16 radials at the
termination end. They are all 130 feet long.

The antenna gets out very well... The capacitor and coil are in a rectangular
Tupperware box attached to a tree with lag bolts.

73,

John, W4NU
Atlanta
__________________________________
hello Jack,

    I use a both a vaccum capacitor and or a regular air capacitor. these work
good for me.I used my 1/4 wave inverted "L" this weekend all the way from
1.801 to 1.865 looking for cw qso's in the contest. If you get any
responses on the torrid, let me know, i'd like to try one of those also.

73 es gl
bill capps
af4od
_________________________

Jack,  The best solution, I feel, is to make the antenna flat top a bit longer
and drive up the impedance closer to 50 ohms.  However this will introduce
reactance that must be canceled by inserting a capacitor.  If it is easy to
adjust the length of the flat top. You might want to start with about 500pf 
fixed
mica capacitor and play with the flat top portion till your VSWR is what you
want. If  you have a good variable that is even better. Be careful with TV 
door
knobs since some cannot handle much current.  I use a 500 pf vacuum variable 
at
home and ended up with about 450 pf to resonate a 65 foot flat top and a 75 
foot
vertical wire.  I have 8 elevated verticals that are grounded only at the 
tower
base.  The vertical wire is about 12 inches from the face of the tower.

Another suggestion is to fold back the flat top if you don't have room for a
longer flat top.  In my case I use a T top and try to get equal amounts of 
wire
on both sides to minimize radiation.  The T can be a true center fed flat top 
or
slightly drooping as in an inverted vee.  However if the angle of declivity 
is to
acute, more than 45 degrees, you will need to add more wire to compensate for 
the
cancellation.

If your feed point is pure resistive,  from 25 to 35 ohms, you can bring down 
the
VSWR by running two pieces of RG-8 in parallel of exactly the same length to 
the
rig. However, the feed line losses in running a 50 ohms feed to a 35 ohm feed
point are small if you can get the rig to tune it.

You can also change the length of the feed line by adding pieces in 25 foot
increments and see if that reduces the VSWR at resonance.  Or another 
possibility
is feeding the antenna with RG-11 or RG6
(75 ohm cable and make the last 1/4 wave of feed line even rolling it in a 
coil
is fine) RG-8 or 213 50 ohm coax.  That 1\4 wave in coax is calculated by
considering the velocity factor of the coax.  If it is the old style RG-8 I
believe the velocity factor is .66 and a bit higher for foam coax.  This
arrangement will work fine if you can coil up the spare coax since the 
additional
losses in feed line is negligible.


Good Luck with you project and let me know if I have been of any help.

73

Herb (ex- KV4FZ)

Good luck and let me know what you end up with

____________________________

Hi Jack ..

The narrow SWR is good, showing that the ground radial system is effective. 
This is the same problem that I have had with the installation here and at 
the 
farm with the full size vertical.

I simply put a 1.78:1 balun in at the feedpoint. I used an Amidon commercial 
unit available from Bytemark (bytemark.com) but one could be wound fairly 
easily.

The ARRL Antenna Book and Handbook should have the designs for this.

I had the same 2:1 SWR and the balun lowered it to 1:1.

Good luck,
Charley W1TE

_____________________________

Loss on 160 from a 2:1 SWR probably isn't worth the trouble of setting up a
matching network at the antenna, unless you have a tremendously long run of
high-loss coax.  For example, with 100 ft of RG-213 and a 25-ohm load, for
an SWR of 2:1 at resonance, the total loss in the coax is 0.276 dB, of
which mismatch loss contributes 0.038 dB.  Your potential loss savings from
a matching network at the antenna amount to something less than 1 percent
of your power, before you figure in the losses in the network.

Both SWR loss and total loss increase some with departures from the
resonant frequency, but that pretty much happens even if the antenna starts
out matched.

By the way, if you want to work through these figures yourself, I generated
them using TLA, which comes with the ARRL Antenna Book.

73, Pete N4ZR
Contesting is!

_______________________________

Hi Jack,

if your antenna is 35 Ohm the SWR should be about 1:1.5 somewere in the 
bandand not 1:2.
You're probably closer to 25/30 Ohm and what you find as a dip at 1830 
is the combination of a reactive impedance transformed by the feed line.
I suggest you to measure at least SWR at antenna base or better to 
measure the impedance with some analyzer.
If the resonance is already close to 1830, in my opinion the best is to 
use a coaxial transformer (1/4 wl line of proper impedance) that's 
inherently good and doesn't require any box or weatherproofing box.
If the resonance is higher than 1.95 Mhz a coil used as an 
autotransformer/fine-loading is the easyest and effective solution.
In any case an L with a coil and a capacitor at antenna base can do the 
job, althought it's mechanical more complex and requires 
weatherproofing.  

73,
Mauri I4JMY

_________________________________

Jack:

A Unun can be wound top match 50-ohms to 34.7-ohms, and in fact 
Bytemark has several ratios available from Amidon if you don't want 
to wind it yourself.
http://www.bytemark.com/amidon/prunun.html 

Winding information is in ON4UN's book, or in Jerry Sevick's 
book "Building and Using Baluns and Ununs". ARRL or Bytemark
have the book. Sevick had information in a number of articles
in QST in the 70s and early 80s, in his excellent series on
short vertical antennas. 
http://www.bytemark.com/amidon/prbooks.html

When I was playing with my inv-L and wasn't sure what ratio I was 
going to require, I purchased the Amidon model W2FMI-HMMU50, 
which allows you to match 50-ohms to five output impedances:
34.7, 22.2, 12.5, 8.68, and 5.56. It handles a full KW, and I left
it at the base of the antenna through a Winter before I replaced it
with a dedicated Unun. Bytemark sells it for $70. This unun is 
documented in Sevick's book as well. 

73,
 Jeff Maass       jmaass@columbus.rr.com

________________________________

If you can lengthen the horizontal portion 50 feet in some fashion.
And use aprox 200pf in series that can handle 10-15 amps of RF current
and a few KV.  You'll end up with a 3/8 wave design that correctly
mates to 50 ohm or 75 ohm coax.  Just adjust the length for resonance
leaving the cap fixed.  Once your resonate under this condition you can
use a variable unit of cap to QSY somewhat if you wish to.  

The advantage here is you get some gain over a quarter wave.  And a cap
will surely have less loss than a transformer.  If you resonate it low
say 1840khz.  And short the cap out with a jumper/relay. You'll end up
with a 3/4 wave resonance on the high end of 75m, with  c. 4db gain or
so.  Leaving the cap in circuit you can use a tuner to encourage 40m
operation on it's near 3rd harmonic, again with gain. BTW the VSWR on
40m will be ruffly less than 2.5 to 1 so you should be still able to load
into it with the amp.  It shows inductive reactance on 40m so you just
need to crank in more C to allow a suitable and acceptable match. 
Another
advantage with this antenna is it isn't very critical to feedline length
on any of these bands.  If you want to get lower on 75m or 80m you'll
need a base loading coil, which you could opt for a mix 2  t-400 and
wind up with ease.   If you use a fixed cap value ei. ceramic
transmitting
doorknobs expect some temp drift but not majorly objectionable. They can
be mounted in length of PVC with end caps and a drain hole or otherwise
to allow them to breeth.  I put mine in a 48in length of pvc pipe with
one end
cap, mounted it on a short mast with a 5deg up orientation with the
feedline
exiting the open end, thermal pipe insulation slid on the feedline keeps
the insects out of the pipe. A few 1/4 inch X 2inch long bolts allow
external access to the radiator and cap connections.

The only precaution for the inductor if used is to keep rain and snow
off it directly.  Once the value needed is determined you pot it in a
PVC end cap with epoxy.  A t-400 is good for at least 1000 watts
with alil overhead, and should do fine for legal limit PEP.  If you
really worried about it you could opt for a t-400-2D which can
easily handle 2000 watts.

73,    Jerry                                     krr2ak@juno.com

________________________

Jack:

The amount of extra loss due to the SWR on the feed line to your antenna is
miniscule. (For example, let's assume that your feed line is RG-213 and is,
say, 150 feet long. For a 2:1 SWR (25 ohms load) at the antenna end, the TLW
program included with the 19th Edition of "The ARRL Antenna Book" says at
1.83 MHz that the total loss in the RG-213 is only 0.429 dB.

If the match were perfect at the antenna end of the coax, the total loss
would be 0.382 dB, a difference of only 0.05 dB -- not worth worrying about
in other words.

So yes, you can use the transceiver's built-in antenna tuner to keep the
transceiver happy while running barefoot -- in other words, preventing the
transceiver's built-in SWR protection circuitry from being activated to cut
back the output power.

And when you run QRO, you can use the Tune/Load controls on your amplifier
to load it properly into the 2:1 SWR it sees in the shack.

The fact that the SWR changes rapidly with frequency is "good" in that it
shows that the losses in your system are fairly low. The sort of 160-meter
antenna system you are using is very similar to what I used to run for about
7 years when I was N6BV/1 in New Hampshire. There I didn't bother matching
at the antenna either.

73, Dean, N6BV

___________________________

Hi Jack...

I'm not yet using any matching on my 100-foot series-fed, top-loaded
vertical. But if you have a copy of ON4UN's book, there is a photo in there
of the matching coil that K3LR uses (it's in the section describing his
3-element parasitic array, which I'm in the midst of constructing). It
looks like it is about 2 inches long or so, maybe an inch and a half
diameter, and goes between the tower and ground. The feedline center
conductor is tapped up the coil to get the correct match.

73/Jon AA1K

__________________________________

Hello, Jack:

I use an inverted L (50 feet vertical over salt water). The base impedance 
of a 1/4 wave antenna
is 12-13 ohms. I use an Amidon UnUn that converts 50 ohms to 12.5 ohms and 
the SWR at
resonance is 1.1:1 with about 50 kHz bandwidth at the 2:1 marks. Without 
the UnUn, the
lowest SWR was 2:1 at resonance exactly like yours. The high power
unit from Amidon is catalog number 2WFMI-4:1-HCU50 abd lists for $59.95. Worth
its weight in gold.
73, Steve AA4V
Isle of Palms, SC

____________________________

I don't know what instruments you have, so my solution may not work for 
you. Here is how I do it.

I made up a universal L network with a rotary coil and large variable 
capacitor of 1500 pf (small coil, BC type variable) that can be changed 
from the input side to the output side. I connect it to the base of the 
vertical and adjust it until I have zero reflected power. Then I measure 
the values of L and C with my Autek RF 1. I then construct an L network 
coil with #10 wire on a PVC form and use large fixed capacitors to make up 
the required C. Sometimes I put in about 100 pf of variable for final 
tweeking, but is usually not needed. These networks tend to be quite broad, 
provided that there is little or no reactance in the system.

This works every time provided that you have the measurement capability and 
of course, the capacitors.

It is likely that your feeding Z is about 25 ohms. Apparently you cannot 
measure it or you would have quoted a number other than a guess.

You could also construct a 2:1 unbalanced to unbalanced transformer, but 
they are a bit difficult to build. You have to make a tap on the 
winding.You may not have the correct core on hand as most of his are U= 125 
material. Not powdered iron sold for most ununs. Details are in the 
Building and Usings Baluns and UNUNs by W2FMI.

You can also use the program which comes with the ARRL Antenna Book. This 
program will calculate the L network component values for you, but you have 
to start off with an accurate measurement at the feedpoint. If you do not 
have a reasonably accurate antenna analyzer, you will likely have a problem 
getting it just right. I know because before I constructed the little 
universal L, I tried doing all that.....with mixed results and a lot of cut 
and try.

If you cannot measure your feed Z, you could perform calculations for a 
range of assummed values and calculate the component values for each. That 
might get you in the ball park, from which you could cut and try until you 
get a decent match. This method is quite frustrating however. It can be 
done, but takes time and patience.

INK N4OO
Sopchoppy, FL

______________________________

Re Jack, W4NF's inverted vee:

An EZNEC model of the antenna with a good ground system like Jack's
yields a feedpoint resistance of about 25 ohms.  This agrees with his
observation of a 2:1 SWR at resonance on a 50-ohm line.

If the transmitter pi-network output can tune into a 75-ohm load, then an
easy method to lower the SWR at the transmitter would be to use a
1/4-wave length of 50-ohm line between the antenna feedpoint and the
transmitter.  Such an arrangement would yield a 75-ohm SWR at the
transmitter of about 1.3:1 at resonance and the 2:1 SWR bandwidth  would
be from about 1.800 MHz to 1.870 MHz.

If the main object is to get a low SWR when running barefoot (i.e., at
the transeiver rather than the amplifier), be sure to include the total
length of coax from the tranceiver to the antenna (via the amplifier).

If the distance to the antenna is greater than 1/4-wavelength of coax
(about 89 feet for poly, 110 feet for foam), then 3/4-wavelengths will
yield the same results.

73, de Earl, K6SE

_____________________________

HI:  MY INVERTED ELL IS ABOUT 170 FT OVERALL WITH 55 FT VERTICAL.
IT IS FED AGAINST AN ASORTMENT OF RADIALS.  WITH A SERIES CAP
OF ABOUT 220 PF IT IS A PERFECT 50 OHM IMPEDANCE. MAKING IT
LONGER THAN 1/4 WAVE GETS THE CURRENT MAX POINT UP HIGHER.
THE SERIES CAP IS MOTOR DRIVEN VAC VAR, SO CAN MOVE UP AND
DOWN THE BAND ALTHOUGH IT IS SELDOM NECESSARY BETWEEN
1860 AND  1820.  GOOD LUCK 73   CHUCK K6RK  PS IT WORKS!

_______________________

dr OM
further to your mail on your antenn
would it be possible if you could forward your VSWR versus freq results if 
possible please.

I would like to compare it with other results I have .

I used a simple toroid transformer in your situation to bring the 50 ohms 
down to something a lot lower.

best 73s

Bob Parkes 
G3REP

__________________________________

W4NF:
>My lowest VSWR is 2:1 at 1.830 and it climbs quickly either side
>of that freq.  I figure that the antenna is probably at about 35 ohms.

>looking for some sort of homebrew design of a matching network

First, it's 25 ohms. And a simple shunt fed inductor at the feedpoint is
very simple, easy and cheap to construct and adjust there. Also known as
a Z match, it connects across the feedpoint with the inner conductor of
the coax at its top. Just trim for 1 to 1 SWR. Ball park is maybe 5
turns of 3 inch diameter air core or a few turns on a T-200 toroid core.

73,     Roy         K6XK/0

_______________________________________

> If the transmitter pi-network output can tune into a 75-ohm load, then
> an easy method to lower the SWR at the transmitter would be to use a
> 1/4-wave length of 50-ohm line between the antenna feedpoint and the
> transmitter.  Such an arrangement would yield a 75-ohm SWR at the
> transmitter of about 1.3:1 at resonance and the 2:1 SWR bandwidth 
> would be from about 1.800 MHz to 1.870 MHz.

Since the systems we use are all normalized to 50 ohms, the 
SWR on a length of 50 ohm cable is constant along its entire 
length except for loss. The SWR, normalized to 50 ohms, would be 
2:1 no matter how long the cable is (ignoring losses, which we can 
do on 160). 

A 50 ohm radio, which virtually all radios are designed to be,  would 
still see a 2:1 SWR normalized to its output impedance. The worse 
thing you can do to a rig is underload the output by terminating it 
with a higher-than-design impedance. 

Correct matching using transmission lines would occur with a pair 
of 75 ohm 1/4 wl lines in parallel. That would give a line impedance 
of 37.5 ohms, with a line-to-antenna SWR of 1.5:1 normalized to 
37.5 ohms. The transmitter-end SWR would be the same, but 
impedance is inverted (1/4 wl away) to 56.25 ohms. The 50 ohm 
normalized SWR would be 1.13 :1, providing an excellent match.

There are other transmission line transformers that will work, but 
the pair-of-parallel 75 ohm lines is the most simple and has 
excellent bandwidth. Coax is also readily available. 

Other than transmission lines, a 2:1 impedance ratio transformer 
could be used or an "L" network. Any of the three methods are 
good.

 
73, Tom W8JI
w8ji@contesting.com

_________________________________

K6XK/0:
> First, it's 25 ohms. And a simple shunt fed inductor at the feedpoint
> is very simple, easy and cheap to construct and adjust there. Also
> known as a Z match, it connects across the feedpoint with the inner
> conductor of the coax at its top. Just trim for 1 to 1 SWR. Ball park
> is maybe 5 turns of 3 inch diameter air core or a few turns on a T-200
> toroid core.

Hi Roy, 

That only works with a unity match when the antenna is shorter 
than 1/4 wl long, so it presents capacitive reactance at the antenna 
terminals.

The antennas capacitive reactance forms the "C" part of an L 
network (series C, shunt L), with the impedance step-down towards 
the antenna.

Unless he is willing to shorten the antenna, he will need to use a 
series capacitor of 3500 pF between the antenna and a shunt coil 
of 4.35 uH across the feedline. Loss will be 0.5% with an inductor 
Q of 200 and a normal capacitor.

He could invert the sign of the components and use a series 
inductance (or make the antenna longer) and use a shunt 
capacitance across the feedline. The series inductance would be 
2.17 uH  and the shunt capacitance (across the feedline) would be 
1740 pF. If the same Q components are used the "no free lunch" 
rule kicks in and loss is again the same 0.5%!!

How we match the system would be mostly a matter of what we 
have laying around, or can get easily. The end result is always 
about the same, the is no free lunch or magic.

73, Tom W8JI

_____________________________

Jack,

N4ZR really hit the correct answer on the head in regards to line loss on 160 
meters and the slight mismatch you have.

73's Don Kirk (wd8dsb)

________________________

Jack:
     May I suggest the setup I use here for my elevated vertical.  It uses a 
100
ft rise and 55 ft run.  The matching network is a series resonate capacitor 
and
inductor. The inductor is ground end and capacitor is antenna end.  Adjust 
coil
and cap for resonate, and tap the inductor with center of coax for best match.
The shield of coax connects to the ground end of coil.  this has worked very 
well
for me here.  If fact I have small DC gearmotor attached to cap to move 
anywhere
in band with good SWR.

73 de Dennis,  w8nw

_________________________

Jack

At 2:1 VSWR more like 25 ohms  so a 2:1 torrid xfrmr is a good bet. Picking 
the right core (material and size) is the trick. W2FMI's book "Transmission 
Line Transformers" is a good read on the subject. At very high power levels 
cores can saturate, go non-linear, create TVI etc, but good designs work 
reliably.

Making the Horizontal leg longer (about 80-90 feet) will raise the R 
component to near 50 ohms, then the matching network becomes a series C of 
about 300pf, makes matching and tuning easy if the C is a variable.

L/C matching networks also work well and allow you to retune the antenna 
for different parts of the band depending on what is going on.

The advantage of the last two methods is that critical tuning can be done 
with our raising and lowering to prune the antenna. I currently use method 
number 2 above but have done all three successfully.

Dave

NR1DX
__________________

Now I have a lot of reading to do and figure out what I have in my junk box 
that I can use.  Thanks for all the good ideas.  CU all on Top Band.  73, 
Jack W4NF


--
FAQ on WWW:               http://www.contesting.com/FAQ/topband
Submissions:              topband@contesting.com
Administrative requests:  topband-REQUEST@contesting.com
Problems:                 owner-topband@contesting.com