Saturday, 26 March 2016

Notes from the Shack - March 2016

It's been a while since I wrote one of these (and I'm sure I've written those same words before). However, it is time to improve the antenna situation at the VE3NRT QTH which will prompt many investigations and conversations about all things antenna. Brad VE3HII stated once that the budget for antennas should be at least 3 times that of the transceiver, and now that I've started budgeting it may turn out to be an understatement.

Up to now there are 4 antennas in rotation at the QTH:
  1. A Butternut HF6V. This is been my mainstay antenna for over 30 years (including 20 years where it sat on the ground beside the driveway at the old QTH in London while I was getting established in the Greater Toronto Area. It is now quite unreliable despite several attempts to refurbish it.
  2. A G5RV at about 35'. Well the middle is about 35'. One end is almost as high. With the recent ice storm, the other end is at about 20'. It still works pretty well and has yielded me quite a number of new DX contacts. The biggest problem with the G5RV is the 150' of not very good coax that feeds it. On 6 metres it's absorbing at least 40% of my output power.
  3. A 40m dipole. I used this at home on 40 and 15 (and sometimes 6) before I had the G5RV. I originally built it on vacation in Florida to run as an inverted V. Consequently it is a bit short to use as a dipole, as inch for inch, an inverted V tunes lower.
  4. A home-made 6m loop. This antenna worked well (Venezuela was my best contact on it) even though it was hung in a tree at less than 20'. It had a tendency to fall out of the tree in a strong wind.
Usually at least one antenna was working at any given moment. But it's time to do something better. The first step in my study has been to look at transmission lines, and the book "Low Band DXing" by ON4UN has been one of the main references. In the chapter on transmission lines, it mentions calculating the loss in coax by measuring the SWR. Now with new coax, you can just look up the loss. Each brand usually has a chart, table, and/or formula for calculating the loss at various frequencies.

However, it might be the case that you have existing coax (whether it be in storage or actual operation) that you want to test. I happened to have a nice 50' piece of RG-213 lying around that I could experiment with, and ON4UN's book had a formula for loss based on measuring VWSR with one end of the cable unterminated. The idea is that you measure VWSR (with an SWR meter or an analyzer - I used the latter) and use the formula to derive the loss.

The formula given in the book is:
$$Loss(dB) = log_{10}\bigg[{SWR+1 \over SWR-1}\bigg]$$
Armed with this formula I measured the cable from 50 to 600 MHz. There were peaks and valleys in the reading (the book never mentioned those) so I used the peaks and came up with an answer of about .4 dB at 570MHz. I realized there must be a mistake as my piece of RG213 had not taken on magical low-loss properties.

In addition, most formulas involving decibels (except ones with ratios of decibels) have the number 10 in them someone, and my result seemed about 1/10th of what it should have been.
Sure enough, on checking some other references I found that the formula should have been expressed as:
$$Loss(dB) = 10 \times log_{10}\bigg[{SWR+1 \over SWR-1}\bigg]$$
Increasing the resolution of the analyzer and reducing the sweep to 550MHz±50MHz showed an undulating pattern of VSWR rising and falling across the range of frequencies. The gap between peaks was about 6.2MHz. An Anritsu web page recommended averaging the SWR in the peaks and valleys. Here's what a 10Mhz segment looks like on the analyzer.


Frequency is on the x-axis and VSWR on the y-axis.

In the 550±50Mhz range the maximum VSWR was 3.525 and the minimum 1.762, yielding an average VSWR of 2.643 and a calculated loss of 3.459 dB. The chart says reads about 6.8 dB so this tells me that the measurements and calculations are likely correct and that the cable is in good condition. A 2005 QEX article by AI1H recommends combining this measurement with the same method applied to a shorted, rather than unterminated, coax. I haven't tried this yet.

The label says that it is JSC Wire & Cable RG213/U part #3780 with a velocity factor of .66. Loss at 400MHz is supposed to be only 4.8dB per 100' at 400Mhz and 8.2dB per 100' at 900 MHz so either the manufacturer is a bit optimistic (say it ain't so), the cable is longer than I think it is, or something else is wrong. So it appears the numbers work well but I want to try a few more examples before declaring that I really understand it.

The above equation means that as the cable loss goes up, the VSWR at the radio end goes down. So cheap cable, especially a long run of cheap cable, will make your VSWR look good but not achieve any real value in terms of communications. Quite the opposite in fact.

A final thought is that the equation for "Return Loss", which is a term I'm starting to notice quite often, is:
$$Return loss(dB) = 20 \times log_{10}\bigg[{SWR+1 \over SWR-1}\bigg]$$
This is just double the value of the cable loss, and represents the "there and back" loss, not just the one-way loss from antenna to transceiver, or vice-versa.

Next steps: measure the cable physically and electrically.

No comments:

Post a comment