Richard Clark wrote:
On Wed, 01 Dec 2004 10:17:05 -0600, Cecil Moore
wrote:

As you know, furnishing line losses in dB implies a power ratio.

dB is dimensionless. Such generalizations forced into logic become
naive paradoxes:

Note from the matched line example above where 5 watts of line losses are
0.21 dB to the mismatched line example where 5 watts of line losses are
1.25 dB (for the same forward power of 100w), the same magnitude of loss
appears much higher as a dB value referenced to NET source power.

with boundary conditions being violated with the substitution of
Mismatch Loss for Dissipative Loss - and done poorly too.

Well, I think he's assuming dissipative loss in both cases. But the 5
watt numbers were assumed, and there's nothing to indicate that the same
piece of transmission line would produce these losses under the
described conditions. Seems likely that the line losing 33% would have
to be quite a bit lossier. As for the second case, we don't know what
the source power was, so it presumably could also be a matched case at
lower power. The example, unfortunately, compared apples to oranges.

My take on dB attenuation figures in wire and cable table data:
dB loss = 10*log(power dissipated in the line / (power dissipated in the
line + power dissipated at load))

By way of reference, according to an old Standard Wire and Cable data
book of mine, the attenuation factor in dB per hundred feet is:

A=4.35*(Rsubt/Zo) + 2.78*sqrt(E)*p*F

Rsubt = .1*(1/d + 1/D)*sqrt(F) (total line resistance in ohms per 1000
ft) [*Note: there might be a misprint here. The .1 would seem to
indicate a conversion to ohms per hundred feet.]

E = dielectric constant

p = power factor of the dielectric

F = frequency in MHz

d = outside dia. center conductor [no units are indicated]

D = inside dia. outer conductor [no units are indicated]

The manual also discusses attenuation as a funtion of VSWR. But they do
not imply that the additional attenuation is due to anything but
reflection. In other words, there is no indication that reflected
'power' causes additional dissipative losses.


And.....I think Reg makes a good point about reflected power.
_____
$0.02

AC6XG