"Roy Lewallen" wrote
You can also determine the loss by measuring the input impedance or SWR
of an open or short circuited piece of cable -- the better the SWR, the
higher the loss. Loss in dB = 10 * log[(SWR + 1) / (SWR - 1)].

===============================

The above formula may be correct but it's not of any practical use.

(1) It applies only when SWR is measured by running alongside a slotted
line with a voltmeter, stopping to make a note of the voltmeter readings at
the max and min points, and then calculating Vmax/Vmin. It is of academic
interest and is entirely unrelated to line Zo.

(2) The error in application arises from the confusion caused by the
old-wives incorrect assumption that the TLI (Transmitter Loading Indicator)
actually measures SWR. It doesn't.

(3) If the ordinary TLI did measure SWR then the measurement would apply
only to 50-ohm lines. On any other line, such as old coax when Zo is
unknown, the measurement would be meaningless.

(4) If the cable to be checked was really 50 ohms and the so-called SWR
meter was standardised on 75 ohms then the measurement would be equally
meaningless.

(5) If the shack 50-ohm SWR meter can be relied upon (and most of them
can't) then the ONLY information to be gleaned is to measure both the
short-circuit and open-circuit SWRs. If they are in the same ball park then
the unknown cable is probably not far from 50 ohms.

(6) Line attenuation is a very small quantity. It is the most difficult of
all line parameters to measure accurately. The last instrument to choose is
the SWR meter in the shack. The most attrocious mechanical damage can be
caused over a short length without hardly any effect on attenuation over a
longer length. To detect the presence of other than catastrophic cable
faults is not possible by attenuation tests. There are far more sensitive
ways of detecting and locating faults.

(7) To check an old length of line just test for DC continuity and look at
the ends. If all is bright and clean there's nothing wrong with it.
----
Reg, G4FGQ