In article Paul W. Schleck " writes:


Reading around on the subject, it seems that there are different kinds
of losses in connectors and transmission lines, and not all of them will
cause heating at the junction or interval where the loss occurs. Some
knowledgeable amateurs have noted that we sometimes confuse transmission
losses and insertion losses. Transmission loss, either Ohmic resistance
or dielectric, will be dissipative, and directly heat the transmission
line where the losses occur. Insertion loss, such as from impedance
mismatch, results in power being reflected. Even impedance-matched
connector conversions (50 Ohm to 50 Ohm) may still result in some
impedance mismatch due to practical limits on their physical design,
including their actual non-constant impedance that can vary with
frequency, and the quality of their connections, particularly after
long-term exposure to the elements. This is where the imaginary part of
complex impedance, also known as reactance, comes in.


The question being addressed was the losses of an adapter when connecting
a cable to an HF rig built with a UHF connector.

The adapters are very short sections of transmission line. The Smith
chart works here, too. Even with a substantial impedance mismatch through
the connector, the segment is very short with respect to a wavelength, and
will not provide a significant input impedance mismatch, if the output side
of the connector is connected to a 50 ohm load.



5. Insertion loss for RF connector adaptors

"1/4 to 1/2 db per connector is the loss I've always heard."

http://www.amsat.org/amsat/archive/a.../msg00214.html


I suggest measurement is a better way to determine the truth. These
claims are easy to test --- measurements I made some years ago didn't
support such claims. In fact, if you are getting such substantial losses,
it is probably time to investigate other possible causes.

Alan