[Antennas] From transmission lines to tuners

2002年12月19日 10:26:48

At 09:30 AM 12/19/2002 -0600, Sandy and Kees Talen wrote:
>Since it's really an antenna "system", as has been pointed out by 
>several, here is a highly recommended article on antenna tuners 
>and their losses. The only thing I walk away with is "where do I
>find the parts to build a Johnson Matchbox equivalent". The next
>thing is how to test your beam (traps/joints, etc) on ALL the 
>elements (and not with an ohm meter). Maybe this has been 
>covered earlier.

I hope I am include in the "several."
>>Frank Witt's (AI1H) articles in 1996 QST are most informative 
>and certainly indicate you can quickly loose more at the tuner 
>than at the transmission line. It also makes a very good case for 
>good link coupled tuners like the Johnson Matchbox vs the 
>many, more compact, tuners we see today. I can see many of the 
>old timers nodding their heads. 

Nod nod. My concerns about tuner losses predate the publication of Witt's
papers. I sent correspondence to ARRL about this in early 1994, however,
it wasn't published.
>>I gather from it, that typical toroid 4:1 balun losses for approx 
>1.5:1 or better SWR are on the order of 0.5dB. Higher losses 
>on the lower 80m band (1dB ....worse on 160m) are due to 
>toroid material selection limitations (selected for good 10m, 
>15m performance, can't cover 10m-160m with one type). 
>>On the Heathkit tuner (similar to many being used today), I can 
>see why lowest balanced line losses are observed using a 200 
>ohm load with a 1:1 SWR due to the 4:1 toroid balun. Why are
>the lowest unbalanced line losses observed using a 200 ohm
>load with a 4:1 SWR ? A 50 ohm load with a 1:1 SWR is also
>low, but the 200 ohm load is lower according to his data ??

Go to:
http://users.triconet.org/wesandlinda/ladder.htm
and read the "Letter to Dean" link and look at the data in the table.
There you will see that *the same tuner* can have different loss on *the
same load* depending on how it is adjusted.
For those that didn't follow the link, read on. For those that didn't
follow the link and think that theory doesn't count, you're excused :)
With decent components in the tuner (not all of them have them) the
inductor will generally be the biggest contributor to insertion loss. This
results from the fact that the unloaded Q of most capacitors is much
greater that that of the inductor. 
Absent voltage breakdown in the components, it is the increased circulating
current in the network that is the major contributor to loss. With a give
unloaded Q, the loss will increase when the loaded Q increases. With a
network comprised of three variable elements there are in theory an
infinite number of settings that will effect a match between two
impedances. Each of them will have a different operating Q, so the object
should be to adjust the network for a match with the lowest network Q.
Without knowledge of the best initial settings *or an output indicator in
the line*, it is easy to get a match at the input of the tuner with less
than optimum insertion loss.
Similarly, to answer the question above, it is likely that a 200 ohm load
results in a reduced loaded Q compared to a 50 ohm load. Therefore, the
efficiency is higher.
This is something that the ladder line proponents seem to forget about.
They get a match to some wildly varying load and figure that the system
must be efficient because everybody *knows* that ladder line is low loss.
On one band, or with one particular line length, where the feedpoint
impedance is "optimum" a given tuner might be very efficient. On another
band, or with a different line length, with a different impedance *but even
with the same SWR*, the efficiency might be terrible.
The problem with the Johnson Matchboxes is that they have limited matching
range. Someone (I believe W8JI) wrote on this before. Also, they are very
difficult to model, which something I highly recommend.
It never ceases to amaze me that with all of the computer tools we have
today, few take advantage of them and instead continue to propagate old
wives tales about what is a "good" system.
Dan, AC6LA, (www.qsl.net/ac6la) has a nice Excel spreadsheet that will
calculate several matching networks, allow them to be "tuned" and will show
the network losses. There are built-in data tables for common transmission
line types or you can specify your own. He also has an antenna modeling
program that uses the NEC-2 engine, or will interface with all of the
various commercial and freeware programs. 
And the best part is that all of this stuff is free! Fewer than ten years
ago, my company spent thousands of bucks for software such as this.
Using just these two free programs, you can model your antenna, get the
impedance data, input it into the spreadsheet, add your favorite
transmission line and calculate a matching network, and see the overall
efficiency.
Wes N7WS
Happy holidays to all.