[TenTec] SWR is a DUMB Indicator of Antenna Performance

Steve Hunt steve at karinya.net
Mon Sep 15 17:54:11 EDT 2014


At HF the losses in most transmission lines are copper losses, not 
dielectric losses; so the losses are directly proportional to the square 
of the current.

Picture the current standing wave along a mismatched TL with an SWR of, 
say, 4:1. At the current maxima the current will be twice what it would 
be in the matched case for the same power, and at the current minima it 
will be half what it would be in the matched case. So the cable losses 
at the current maxima will be 4 times what they would be in the matched 
case, and at the current minima they will be one quarter what they would 
be in the matched case.

Integrated over a long length of line the net effect is more loss than 
in the matched case; but for short lengths of line that isn't 
necessarily true.

Take as an example calculating the losses in 2ft of RG58 wound on a 
ferrite core to form a 1:1 balun. Feeding a matched 50 Ohm load with 
100W the differential-mode current in the coax will be 1.4A; But if we 
now swap the 50 Ohm load for a 200 Ohm load (SWR=4:1), the current at 
the load end of the coax will drop to 0.7A and the power loss at that 
point in the cable will be one quarter what it was in the matched case. 
Further back from the load, the current in the cable will rise due to 
the standing wave pattern, and the difference reduces; however for short 
lengths of cable the integrated loss can be less than the matched loss.

At, say, 21MHz the matched loss for 2ft of RG58 is 0.036dB, but with a 
200 Ohm termination it drops to 0.022dB - about 40% less.

The converse happens if the termination is a low resistance, because the 
current is then higher than in the matched case. With a 12.5 Ohm load 
the loss rises to 0.131dB - significantly more that in the matched case.

So there we have an example where the cable losses for the same 4:1 SWR 
are very different - in one case higher than the matched loss and in the 
other case lower than the matched loss. Therefore any loss calculator 
which asks for SWR as an input parameter rather than actual load 
impedance is making assumptions which might not be valid! And any table 
that simply shows "additional loss" due to some particular SWR should be 
treated with caution.

Hope that's clear,
Steve G3TXQ




On 15/09/2014 22:13, Bob McGraw - K4TAX wrote:
> OK, I'll fall for it, how does a cable have a negative dielectric loss 
> figure?
>
> 73
> Bob, K4TAX
>



More information about the TenTec mailing list