I'm not sure it helps to distinguish between "SWR losses" and
"I-squared-R losses" - at HF, all the losses are predominantly
It may help to picture qualitatively the current profile over a short
length of feedline at the load end - in all cases delivering the same
power to the load:
* If we have a matched load, the current is constant along the line and
the loss-per-unit-length will therefore also constant along the line.
Cumulative losses increase linearly with length.
* If we have a load with a moderately high resistive component the
current at the load will be lower, and therefore loss-per-unit-length
will be lower. But slightly back from the load the current will have
increased due to the standing wave pattern, and therefore the
loss-per-unit-length will be higher; eventually it exceeds the matched
case loss-per-unit-length, and even further back the cumulative losses
exceed those of the matched case.
* If we now have a load with a *very* high resistive component, the
current at the load will be very low and the loss-per-unit-length will
be even lower than in the previous case. However, moving back from the
load, the rate of change of current with distance is higher because of
the increased ISWR, and it may be that we reach the "break even" point
sooner, despite the loss-per-unit-length adjacent to the load being lower.
Incidentally, there will be a load value which maximises the distance
from the load of the "break even" point.
So, all the losses are "I-squared-R losses" - it's just that the current
profile (and therefore the cumulative loss profile) changes if the ISWR
is not unity.
On 01/12/2010 13:44, Paul Christensen wrote:
> The additional loss attributed to a mismatch is still relevant once SWR
> becomes part of the loss. In the 10 ft. examples we've been using, that
> occurs when the line get to roughly 40 degrees in length. When we approach
> 1/4 wave, loss due to SWR becomes equal in loss to I-squared-R loss. I
> think that's was one of Steve's points in that the additional loss
> attributed to SWR needs some clarification in the footnotes o be completely
> Paul, W9AC
> TowerTalk mailing list
TowerTalk mailing list