[TowerTalk] Measuring Coax Loss

[Rick Karlquist]

>> > That's my point. If you're measuring a long piece of cable at a
>> > hamfest, the
>> > MFJ is plenty good enough. If you're measuring a 10 ft piece of cable
>> > but
>> > want to know how it would work on a very long run, especially at low
>> > frequencies, you need the greater precision of the HP setup.
>> >
>> > 73,
>> >
>> > Jim K9YC

If I wanted to accurately measure the loss on a 10 ft piece of cable,
I would short circuit one end.  I would put a TEE on my signal
generator front panel connector and connect a power meter sensor
to the side port of the TEE.  Next, I would note the power meter
reading of the sig gen output.  Then, I would then drive the unshorted end
of the 10 ft cable with the third port of the TEE.  The cable loss can be
very accurately calculated by the difference between the power meter
reading with vs without the coax under test attached.  The lower the power
meter reading, the lower the loss of the cable.  For example, if the loss
of the 10 ft cable is 1% (about 0.043 dB), the power meter will read
28 dB lower when loaded by the coax than unloaded.

You don't need any heroically accurate instruments to make this
measurement.  You could almost do it using the S-meter on a radio if you
previously checked the S-meter characteristics.  A 3 dB error in the dB
measurement only represents a 10% error in the cable loss, due
to the way the math works.  You can also use a variable attenuator
on the sig gen to make the measurement.  In that case, the
S-meter is only used at one setting and its calibration is irrelevant
(this is a well known technique).

Rick N6RK