[TowerTalk] Double Coax Runs to Reduce Loss?

[jimlux]

On 12/18/16 9:48 AM, rfman45 wrote:
> I was at a hamfest recently and a seller there was discussing how he uses two equal length runs of the same coax (an LMR type) to halve his losses on a long run to a distant tower some 300 or so feet away. His point was that he had a lot of this type of coax around and therefore it was better to do this than to buy something more efficient.
>

How does he do impedance matching?

But, in general, no, this doesn't help - Consider the loss of a length 
of coax as being, say, X dB.  Now split your power into two paths, and 
send each half down a separate piece of coax.  You'll get X dB loss in 
each path, and when you combine them at the other end, you'll have X dB 
total loss.

There are some cases:  where the lines are mismatched, or very short - 
the "X dB/100 meters" is for a "long transmission line" where theres 
multiple wavelengths.  If you've got a 1/8th wavelength and the system 
isn't perfectly matched, the actual loss will depend on whether you're 
at a high current point or a high voltage point.  For HF, the dominant 
loss is IR losses not dielectric dissipation.

SO let's think of another case:
Say the resistance of the coax is 1 ohm.  If I transform my signal to 
100 ohms, the current will be halved, the voltage doubled, and run two 
pieces of coax in series (so it's 100 ohms).  But now I have 2 ohms in 
the transmission line, and the loss is exactly the same.