Need help... End-fed, long wire or ????
 

In article t3ali.3392$Y_3.570@trnddc04, Woody wrote:

Well.. a million thanks for that. Quite a cool history lesson as well. So
now I'm looking for an SEA tuner... LOL...

Listen, that all makes perfect sense but just to clarify, a.) now I know why
that triton did so poorly when tested. We calc'd 1/2 wavelength for the
longwire, and b.) Again, for continuity and clarity of this thread for
future surfers...... what then, considering our discussed auto-tuners, would
be the optimal length for a longwire that would be used for amateur/MARS,
3-30MHz?
Pick 1/2wavelength on say 2.8Mhz and just cut it? Or calc 1/2wavelength on
the lowest and add 5% or some arbitrary odd number??
Which plan will offer the least chance of dropping a 1/2wl further up the
band on a desired frequency?

I think you'll need to run a a simple calculation, based on the
frequencies you actually want to use.

What you'll want, is a wire whose length is not particularly close to
any multiple of 1/2 wavelength, on any frequency you want to use. A
wire which would match easily on 80 meters (e.g. 1/4 wavelength long)
would be a bad choice if you want to work on 40 meters as well, as
it'd be 1/2 wavelength long.

A simple program or spreadsheet ought to be able to do the necessary
calculations... try every wire length from 66 feet to 132 feet and see
if you can find a length which is a comfortable percentage away from
an even multiple of 1/2 wavelength on each frequency. Or,
alternatively, iterate through each frequency, calculate the
1/2-wavelength multiples, and "blacklist" every possible length which
is too close to these multiples.

For what it's worth, SGC sells a longwire antenna 60' in length, which
they say works well on both lower and higher HF bands. It might or
might not be a good choice for MARS frequencies.

--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

Need help... End-fed, long wire or ????
 

Bruce in Alaska wrote in
:
Bruce,

An interesting post overall, but ...

System. All these tuners NEED a Low Impedance RF Ground to work
against, as well as a Longwire who's length is SPECIFICALLY set up to
put the 1/2 Wavelength Point in a non used portion of the Spectrum.
They will NOT tune within 2% of the Natural 1/2 Wavelenth point of the
Longwire connected, where Antenna Impedances near Infinity.

The "rules" above are offered without definition or explanation, so they
don't really raise the art.

Manufacturer's "rules" and explanations are often inconsistent. For
example, some state a minimum length that can be tuned on say 3.6MHz, and
it is often around 3m, yet they rabbit on about avoiding half wave
resonances in the wire to avoid high voltage at the feed point. The feed
point voltage on a 3m whip at 3.6MHz is likely to be much higher than a 10m
whip at its first parallel resonance.

Owen

Need help... End-fed, long wire or ????
 

Need help... End-fed, long wire or ????
 

Dave Platt wrote:
In article t3ali.3392$Y_3.570@trnddc04, Woody wrote:


Well.. a million thanks for that. Quite a cool history lesson as well. So
now I'm looking for an SEA tuner... LOL...

Listen, that all makes perfect sense but just to clarify, a.) now I know why
that triton did so poorly when tested. We calc'd 1/2 wavelength for the
longwire, and b.) Again, for continuity and clarity of this thread for
future surfers...... what then, considering our discussed auto-tuners, would
be the optimal length for a longwire that would be used for amateur/MARS,
3-30MHz?
Pick 1/2wavelength on say 2.8Mhz and just cut it? Or calc 1/2wavelength on
the lowest and add 5% or some arbitrary odd number??
Which plan will offer the least chance of dropping a 1/2wl further up the
band on a desired frequency?


I think you'll need to run a a simple calculation, based on the
frequencies you actually want to use.

What you'll want, is a wire whose length is not particularly close to
any multiple of 1/2 wavelength, on any frequency you want to use. A
wire which would match easily on 80 meters (e.g. 1/4 wavelength long)
would be a bad choice if you want to work on 40 meters as well, as
it'd be 1/2 wavelength long.

A simple program or spreadsheet ought to be able to do the necessary
calculations... try every wire length from 66 feet to 132 feet and see
if you can find a length which is a comfortable percentage away from
an even multiple of 1/2 wavelength on each frequency. Or,
alternatively, iterate through each frequency, calculate the
1/2-wavelength multiples, and "blacklist" every possible length which
is too close to these multiples.

For what it's worth, SGC sells a longwire antenna 60' in length, which
they say works well on both lower and higher HF bands. It might or
might not be a good choice for MARS frequencies.


The other strategy is to use two wires of appropriately different
lengths connected together at the feedpoint. (the SGC whips do this, for
instance).. Space the wires some distance apart (a few inches would
do).. What this does is put multiple bumps in the impedance curve and
eliminates the pathological cases where you have very high Z when the
(one) wire is a half wavelength or multiple thereof. At those
frequencies where one wire *is* a half wavelength, and presents a high
Z, the other one is likely NOT a multiple of a half wavelength, and so,
will present a reasonable impedance. Sure, they interact (as folks
making multiband dipoles find when trying to cut and trim), but all that
really does is shift the resonances around.

An interesting question would be what is the optimum ratio of lengths..
probably something like 1:1.618 or 1:2.7183

Need help... End-fed, long wire or ????
 

Owen Duffy wrote:
Bruce in Alaska wrote in
:
Bruce,

An interesting post overall, but ...


System. All these tuners NEED a Low Impedance RF Ground to work
against, as well as a Longwire who's length is SPECIFICALLY set up to
put the 1/2 Wavelength Point in a non used portion of the Spectrum.
They will NOT tune within 2% of the Natural 1/2 Wavelenth point of the
Longwire connected, where Antenna Impedances near Infinity.


The "rules" above are offered without definition or explanation, so they
don't really raise the art.

Manufacturer's "rules" and explanations are often inconsistent. For
example, some state a minimum length that can be tuned on say 3.6MHz, and
it is often around 3m, yet they rabbit on about avoiding half wave
resonances in the wire to avoid high voltage at the feed point. The feed
point voltage on a 3m whip at 3.6MHz is likely to be much higher than a 10m
whip at its first parallel resonance.


there's a goodly amount of empiricism in these rules of thumb from the
mfr, too.

If you start to analyze it, there's all sorts of issues that crop up in
the analysis: the heating and/or breakdown of the Ls and Cs inside the
box, for instance (and, because it's a binary sequence with components
with 20% tolerance, the voltages (for L) or currents (for C) don't
divide evenly... the big Cs take more of the current than the small ones).

And then, there's the whole thermal management issue, and "real ratings"
of the component vs catalog vs derating. The tesla coil folks regularly
run low ESR extended foil polypropylene capacitors (typical 0.15 uF at
2kV) at 2 or 3 times rated voltage at a few hundred kHz, as long as the
rms current is within bounds. But they benefit from a fair amount of
destructive testing and failure analysis by a coiling hobbyist with
access to the appropriate gear.


Owen

Need help... End-fed, long wire or ????
 

In article ,
Owen Duffy wrote:

A simple program or spreadsheet ought to be able to do the necessary
calculations... try every wire length from 66 feet to 132 feet and see
if you can find a length which is a comfortable percentage away from
an even multiple of 1/2 wavelength on each frequency. Or,

You probably mean't any integral multiple of a half wave.

You're right... integral multiple of a half-wavelength, or even
multiples of a quarter-wavelength are two alternative ways for stating
the lengths to be avoided (high feedpoint Z).

alternatively, iterate through each frequency, calculate the
1/2-wavelength multiples, and "blacklist" every possible length which
is too close to these multiples.

I have done just that, and searched for "sweet" wire lengths that aren't
within say 5% of band edge for all HF bands. It sounds like a solution to
the problem doesn't it. (5% implies that you have a pretty determinate
scenario, which is a big assumption. IIRC 10%+ will not give a practical
result on the higher bands.)

Problem is that it probably unecessarily constrains the solution.

Entirely possible!

The input impedance and the feed point voltage of an end fed wire at its
higher parallel resonances falls, so that whilst you might want to avoid
the first such resonance, the impedance (and feed point voltage) at the
third or higher resonance might well be low enough to not worry about it.

And, even if it was a bit high, you might not need to be more than a
couple of percent away from it to get it down to a length that might
work.

Odds are that some amount of experimentation is going to be required,
at any given installation, to find a wire length which tunes up well
with these ATUs. The orientation of the wire (vertical, inverted-L,
etc.), height about ground, presence of trees and metallic objects,
and (perhaps most importantly) the details of the ATU's grounding
system, are likely to change the impedances around enough to make the
"textbook" answers less than completely effective.

--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

Need help... End-fed, long wire or ????
 

Hi Owen. 5% was just a number picked out at random by me for clarification
of my question.
So you wouldn't want to share your 'sweet' findings, would you??
thanks!
W


"Owen Duffy" wrote in message
...




I have done just that, and searched for "sweet" wire lengths that aren't
within say 5% of band edge for all HF bands. It sounds like a solution to
the problem doesn't it. (5% implies that you have a pretty determinate
scenario, which is a big assumption. IIRC 10%+ will not give a practical
result on the higher bands.)
Owen

Need help... End-fed, long wire or ????
 

"Woody" wrote in news:hYdli.29709$BT3.8034@trnddc06:

Hi Owen. 5% was just a number picked out at random by me for
clarification of my question.
So you wouldn't want to share your 'sweet' findings, would you??
thanks!

W,

I have had a quick look for the spreadsheet and Perl scripts, but haven't
found them. I am sure I have them, but haven't filed them in an ordered
way apparently, I think that technically is lost... or galloping
senility.

The reason I didn't publish them as is at the time is is that they are an
incomplete picture.

Have a look at the article at
http://www.vk1od.net/InvertedL/InvertedL.htm which describes an InvertedL
at approximately one of those "sweet" lengths (~26m). In fact, the length
was juggled to avoid excessive feedpoint voltage on all bands.

Another of the "sweet" lengths is half that at 13m, and the voltage plot
is rougly scaled proportionately in frequency.

You will see from Fig 1 that the voltage peaks at higher frequency
parallel resonances are less an issue than the first and second
resonances (you can't see from the graphs, but ~10kV and 3kV
respectively).

Of course, none of this discussion addresses the pattern issues at the
higher frequencies.

As far as the earth system goes, it impacts efficiency of the system. It
is my view that the earth only needs to be good enough that its loss is
an acceptably low portion of the transmitter power. The shorter the
radiator in wavelengths, the lower its Rr and therefore the lower the
earth resistance for comparable efficiency. If you look at Fig 4 of the
article, you will see that Rr is 100 ohms or greater above 5MHz, so the
loss of an earth system resistance of say 30 ohms is near insignificant,
but at 80m where the length is relatively short, you need a better earth
for good efficiency.

Don't agonise over it too much, and treat the Rules of Thumb as ROT until
you understand the underlying assumptions and caveats.

Owen

W


"Owen Duffy" wrote in message
...




I have done just that, and searched for "sweet" wire lengths that
aren't within say 5% of band edge for all HF bands. It sounds like a
solution to the problem doesn't it. (5% implies that you have a
pretty determinate scenario, which is a big assumption. IIRC 10%+
will not give a practical result on the higher bands.)
Owen

Need help... End-fed, long wire or ????
 

"Woody" wrote in news:hYdli.29709$BT3.8034@trnddc06:

Hi Owen. 5% was just a number picked out at random by me for
clarification of my question.
So you wouldn't want to share your 'sweet' findings, would you??
thanks!

I had a dig around through a few thousand spreadsheets... and found it.

9.05m, 12.6m, 18.3m, 26.4m, 33.3m, 44.9m.

If you want to cover 60m, leave the 26.4 out of the list.

Did someone mention 60' (18.3m), it is in the list, but it is more
critical than 26.4 or 12.6.

As noted in my earlier post, if you want to be 10% or more away from n
half waves on all bands, there is no solution.

These calcs don't consider proximity of objects that may shift the
tuning.

My advice still stands that the higher order parallel resonances aren't
such an issue, so this method unnecessarily constrains the solution..

Owen

Need help... End-fed, long wire or ????
 

Yes, the history of the autotuners is fascinating...
The discussion of integral half waves, parallel resonance impedences,
ground impedences, etc., is meat and potatos for antenna freaks like
me...

BTW, in the older ARRL handbooks is a table of guy wire lengths (for
towers) that avoid resonances in the ham bands... Those lengths would
make an excellent guide for the length of a random wire antenna... It
might also be in the new editions, but I haven't looked at one in
recent years.. Anyway, that table will give you a good starting
point...

B U T , what I want you to do is to simply either cobble up an L
match, or pick up a tuner (any tuner, automatic or not), run some wire
any length through those trees, and get on the air!
If the tuner has problems with any particular band, change the length
of the wire by 4 to 12 feet until it tunes OK...
Once you have a working antenna you can spend time and energy gilding
the lily with just the 'perfect' length and configuration...

Don't waste time over analyzing - pick a tuner, throw up some wire
and ground radials, and operate...
A local ham who lives in a trailor park is closing in on 300 countries
confirmed and his only antenna for all this is a single multiband
vertical, 21 feet tall, disguised as a flag pole...

Just do it...

cheers ... denny