Ed wrote in
. 192.196:




I am not sure exactly what you mean by a "a wire antenna, 15 feet
above a flat metal roofed building, (it a NVIS antenna), and resonant
at 6.2 MHz".

I assume you mean a centre fed dipole that is an electrical half wave
at 6.2MHz.


Yes. My appologies for leaving out a critical description. Its a
wire centerfed halfwave antenna, 15 feet above a flat metal roof, and
resonant at 6.2 MHz.

The people who installed it expected the tuner in the FT900 radio
to
handle the matching. I explained that the tuner was never intended
to match an antenna that would be so far off resonance. To compound
matters, they are using at least 100 feet of 50 ohm coax... mostly
RG- 8, I think.



The efficiency of the radiator component can be estimated from an NEC
model, it is probably very high and not the real issue.


Well, in this case, it IS a necessary issue. I wish to present

Whilst neither of us KNOW whether it is an issue, I suggest to you that
based on experience radiator loss in this type of configurationis
probably less than 2%.

the
probably actual feedpoint SWR to these people so they will realize how
far beyond tolerance they are trying to use that built in tuner. That
is why I was asking for SWR estimates for 3.95MHz, 7.2MHz, and
5.4MHz.

You seem to be equating feed point VSWR and that seen by the built in
tuner. That might be the case if the transmission line was lossless. It
is unlikely to be, and we don't know again due to lack of information.

You might find my article at http://www.vk1od.net/LOLL/index.htm which
looks at multi-frequency use of a coax fed centre fed dipole of interest.
Though it models a 66' dipole at 30' height, some (most) of the messages
are relevant to your situation.

Perhaps on reviewing the article, you will see my point that radiator
loss is less important, and that understanding transmission line loss is
very important to your problem.

The article might even suggest some other options that you haven't
nominated.

Owen