[TowerTalk] How lossy are PL-259s at HF?

[Jim Brown]

On 6/17/2024 6:39 AM, Pete Smith N4ZR wrote:
> Just wondering because I am having to redo the feedline from my shack to 
> my tower (about 160 feet), and can avoid buying new coax by using 
> several pieces with PL-259s in between.

There is much junk science and pure BS written about UHF connectors and 
loss. The short answer is as others have advised you -- UHF connectors 
are just fine well into the UHF spectrum, many good engineers consider 
them mechanically superior to alternatives like Type N. Those who think 
otherwise failed (or skipped) their study of Transmission Lines. And 
those measuring loss are either failing in their measurement techniques 
or using junk connectors, of which there are many.

Yes, Zo of UHF connectors is not 50 ohms, but the discontinuity 
introduced by a junction using these connectors is both too small and, 
most important, too short as a fraction of a wavelength, to contribute 
significant loss in an analog system working up to low UHF. Data systems 
would be a very different story -- the discontinuity could contribute 
significant smearing of the waveform.

A data point:  About fifteen years ago, I built a bunch of cables for a 
DXPedition using Commscope 3227, a high quality RG8-sized coax with a 
solid #10 copper center and a shield composed of a dense tinned copper 
braid and high quality foil. The cable's spec STARTS at 500 MHz -- it 
was purchased for use in a data center by a company that went bust 
during the dot.com boom 20+ years ago, WA6NMF bought a truckload of it 
for pennies on the dollar, and I ended up with a dozen spools over the 
years.

The cables were each 100 ft long, and I tested them by wiring 13 of them 
in series and measuring their loss. The connectors were Amphenol 
83-1SPs, and barrels were Amphenol. In those days, my best 
instrumentation was an HP gen and HP spectrum analyzer. The measured 
loss at 200 MHz, where I stopped, was slightly less than the cable spec. 
I started measuring at low enough frequency that I saw the effect of Zo 
being complex and non-constant in that range.

73, Jim K9YC