80 m Dipole Bandwidth

Some recent conversation in our morning group touched upon the bandwidth of an 80 m dipole. We know that the commonly used wire diameters do not allow low-SWR coverage of the entire 500 kHz of bandwidth. Here are model results showing the bandwidth of a theoretical dipole that is fed with a 73-ohm source.

126 ft dipole; no feedline

 This first figure shows the SWR vs frequency for a 126 ft dipole at 50 ft above ground. The source (transmitter) is connected directly to the center of the dipole. Choosing the data points close to the SWR=2 line gives a bandwidth of about 230 kHz (marked in green). At SWR=3 the bandwidth is about 380 kHz (marked in orange).

The next figure is for the same dipole with a feedline having 73-ohm characteristic impedance and fed with a 73-ohm source.

125 ft dipole; 73 ft feedline

The bandwidth for this case at SWR=3 is about 360 kHz -- slightly less that the first case. Additional data points would be needed to more precisely locate the frequencies where SWR exactly equals 3.

An additional model case for feedline of 70.3 ft yielded very close to the same results as for 73 ft. There is little effect from changing the feedline length when operating with feedline characteristic impedance and source impedance close to the impedance of the dipole near resonance.

UPDATE:  I changed the feedline and source (transmitter) impedance to 50 ohms, leaving the other parameters unchanged. The model gave:

• SWR less than 3: bandwidth 280 kHz
• SWR less than 2: bandwidth 150 kHz
• minimum SWR = 1.37

For the cases of 300-ohm or 450-ohm feedline, the SWR is greater than 3 across the entire range of 3.5 to 4.0 MHz, requiring the use of an antenna tuner or other impedance matching methods to operate with a 50-ohm transmitter.


A dipole that is longer or shorter than 1/2 electrical wavelength at the operating frequency will have considerably different SWR than the above ideal cases. Likewise, using a feedline that is much different than the dipole's resonant impedance will affect the SWR significantly. Under these conditions a different definition of bandwidth is needed and relates more to the impedance range that an antenna tuner can accommodate, since the antenna will present a much higher SWR on its own.

WA5MLF

Static Crashes on 3740

Here is an annotated screen shot of the waterfall display showing the static crashes as pale or brighter horizontal strips. The time scale is about 47 seconds from top to bottom.

QRM on 60 m

Here is a screen shot of waterfall display during KE4ID's turn on 60 m. Some Spanish speakers were operating 2 kHz above -- at 5405.5 kHz at 0734 CST.

80 m Loop Feedline Analysis

Recently I used NEC antenna modeling software to analyze the effects of different feedline lengths on SWR for an 80 m horizontal loop at 40 ft height above the "medium hills and forest terrain" ground model. The goal was to see the effects of various feedline lengths that either conformed with or departed from the lengths recommended by W8JI in his article Choosing the Correct Balun in table  3.

I collected the model's data output in increments of 0.2 MHz from 3 MHz to 29 MHz. (The modeling software permits a maximum of 256 data points.) The data were collected for the following lengths of 450-ohm feedline:  384, 400, 418, 467, 512 ft.

The graph below presents the calculated SWR at nine selected frequencies. All feedline cases show a high SWR peak in the 60 m band.  (Click on the image for a larger view.)

 All lengths have similar SWR values at the lower frequencies, and more variation at the higher frequencies. The recommended lengths of 384 and 512 ft (multiples of 127.9 ft) seem to have a bit less variation in SWR at the high frequencies, although 418 ft behaves pretty well. 400 and 467 ft look less favorable.  In the real world, experimentation will lead to an optimal or an acceptable feedline length.  Naturally, it is best to avoid combinations that give very high SWR in a frequency range of interest, since the high SWR will require the antenna tuner to compensate more and the overall antenna efficiency is reduced.

Below are the full calculated SWR plots from the modeling program for each of the feedline lengths. No other antenna parameters were changed. (Click on each image for a larger view. These are reduced-resolution copies, but should suffice for overall impressions.) I found that shorter feedlines, especially 128 and 256 ft (not presented here), gave smoother curves, but I don't have an explanation.  The plots start at 3 MHz, for nicer scaling on the frequency axis, but the antenna is designed for about 3.5 MHz, so the SWR is high at 3 MHz. 

384 ft feedline

400 ft feedline

418 ft feedline

467 ft feedline

512 ft feedline

The program's output numeric data files provide SWR, reflection coefficient, resistance, reactance impedance and phase for every frequency.  Radiation pattern data are also available and can be displayed graphically. For this analysis, the radiation patterns are all identical, since the model's transmission line element is specialized, compared with the wire elements, and does not contribute to the radiation pattern. If the radiation effects of a feedline must be considered in a model, then it must be included in the model as a pair of parallel wires.  For more information on this topic, see part 4 of the QST article series "A Beginner's Guide to Modeling with NEC" by L. B. Cebik.

75 m spectral view at 0738 CDT

Here's a view of 75 m showing low signal levels for this time of day.

Java updates -- suppress sponsor offers

Those who use Java on a regular basis know that when a version update is published, its installer offers to include some third-party programs that are selected by default to install. Yesterday I learned from the Security Now netcast that the Java control panel includes an option, in the advanced tab, to suppress the offering of third-party programs. The screen shot below shows the location of this option.

 
I tested it on a PC this morning and confirmed that it works.

-John

Understanding Linux

The website How-To Geek published this article about 10 popular distributions of Linux. Of the ones listed, I have used, to various degrees, Ubuntu, Mint, Slackware and Puppy. I'm using one as I write this, but the web is (thankfully) agnostic.

The above article contains several links to other articles that can provide a better understanding of what all the Linux choices and piece-parts are about. If you are curious, I suggest browsing the following articles in this order:

What Is A Linux Distro
A brief discussion of the concept of a Linux distribution

8 Pieces of Software That Make Up Linux Systems
This one covers in more detail the components of a distribution.

Linux Desktop Environments
This one covers various desktop environments that may be part of a distribution or may be pulled into a Linux installation. A Linux user may choose from among several distributions that are identical except for the desktop environment. Those who wish to use or try Linux on older hardware are often advised in web articles or postings to consider the light-weight desktop environments Xfce or LXDE. Some Linux veterans have actually set up their installations to enable switching between two different desktop environments. Most users probably try out several distributions, running from a live CD (slow) or a USB flash drive before choosing one or more to install in some permanent fashion.

The web site Distro Watch is my favorite for finding and exploring different distributions. You can browse the main page or use the search filters to find candidate distributions for certain needs. You'll also note that the site catalogs BSD and Solaris, in addition to Linux. For those who are curious, this article explains the difference between BSD and Linux. The use of BSD as the base of Mac OS X is mentioned.

John