On 1/17/12 10:36 AM, Ken wrote:
> On Jan 17, 2012, at 1:29 PM, Jim Lux wrote:
>> Now put that matching network at the feedpoint and see what losses you
> What difference does it make, we don't put the matching network (lump) at the
> feed point (except people with remote tuners like the Icom AH-4). Most of us
> have a linear "matching network" of 30-100' of open wire or ladder line, both
> of which are fairly efficient.
If you're building a 80m single band antenna, and it's short, and you
want to feed with coax, then putting a lumped network at the feedpoint
that gets you "close" to 50 or 75 ohms (depending on your coax) is a
pretty efficient way to work. You can then use the tuner in your shack
(or the tuning on your tube amp) to "trim out" the remaining mismatch if
desired. This is what a "shorty 40" basically does.. shortened elements
loaded with a lumped inductor at the feedpoint, and the inductor also
acts as a autotransformer to get the R up.
Using transmission lines as transformers sometimes works, sometimes
doesn't. They're not lossless, and the tables and a lot of the
calculators make assumptions when calculating "mismatch losses" like
that you've got a "long" length of line. (that is, they're giving you
the average loss). In some cases, a lumped element is going to be lower
loss than a hunk of transmission line, something that folks making stub
tuners and notch filters are very aware of.
If the line is substantially shorter than a wavelength, then the simple
approximations of loss with a mismatch aren't valid.
Take an example of a medium length line (more than half wavelength)
feeding a very mismatched antenna at HF, and look at the loss in each
inch of the line.. At HF, the dominant loss source is IR loss in the
conductors. At some places, the current in the line is high, and you'll
get big losses in that particular inch, but in other places, the voltage
is high, and you get low loss.
If you had a short line (say 1/8 wavelength), feeding a short antenna
(low R, big X), with an opposite X at the sending end of the line, you
could actually have fairly high losses, because the current between the
Xes is high.
TowerTalk mailing list