It's interesting to model this sort of arrangement, note the wide
range of feedpoint impedances that will be seen on the different
bands, and observe the feedline losses that might be incurred when it
is directly fed with 50 ohm coax. If you then introduce an ideal
transformer at the feedpoint and repeat the exercise you will
generally find that feedline losses increase on some bands and
decrease on others.

With the vertical length I tried, the effect of a 1:9 transformer was
to limit the more extreme losses at the cost of making some very low
losses higher. I guess over the several bands I tried you could say
there was a net improvement with the transformer.

But one question would be how to build this ideal 1:9 transformer
which maintains its transformation ratio and exhibits zero loss across
the wide range of impedances and frequencies involved.

Steve G3TXQ

In message
,
steveeh131047 writes
It's interesting to model this sort of arrangement, note the wide
range of feedpoint impedances that will be seen on the different
bands, and observe the feedline losses that might be incurred when it
is directly fed with 50 ohm coax. If you then introduce an ideal
transformer at the feedpoint and repeat the exercise you will
generally find that feedline losses increase on some bands and
decrease on others.

With the vertical length I tried, the effect of a 1:9 transformer was
to limit the more extreme losses at the cost of making some very low
losses higher. I guess over the several bands I tried you could say
there was a net improvement with the transformer.

But one question would be how to build this ideal 1:9 transformer
which maintains its transformation ratio and exhibits zero loss across
the wide range of impedances and frequencies involved.

You might like to refer to the thread "UNUN Cores?How To Wind?",
started on 1 August. In particular, the last posting (by 'UK Monitor')
suggests a link to this website:
http://g8jnj.webs.com/currentprojects.htm

--
Ian

steveeh131047 wrote:
It's interesting to model this sort of arrangement, note the wide
range of feedpoint impedances that will be seen on the different
bands, and observe the feedline losses that might be incurred when it
is directly fed with 50 ohm coax. If you then introduce an ideal
transformer at the feedpoint and repeat the exercise you will
generally find that feedline losses increase on some bands and
decrease on others.

With the vertical length I tried, the effect of a 1:9 transformer was
to limit the more extreme losses at the cost of making some very low
losses higher. I guess over the several bands I tried you could say
there was a net improvement with the transformer.

But one question would be how to build this ideal 1:9 transformer
which maintains its transformation ratio and exhibits zero loss across
the wide range of impedances and frequencies involved.

Steve G3TXQ

Some time ago I made a series of careful measurements of a transformer
which was at the feedpoint of a multiple band antenna. At frequencies
where the feedpoint impedance was very much different from the (purely
resistive) design impedance, the transformation wasn't equal to the
design transformation, and the transformer introduced both series and
shunt impedance. At some frequencies, these effects were extreme, and
the transformer acted nothing at all like an ideal transformer.

Modeling a system like this with an ideal transformer might be an
interesting intellectual exercise. But that's all it is -- the real
system won't behave anything like the model. You can extend a
transformer's range of impedances and frequencies by using great care in
the initial design and construction, then adding compensating circuitry.
The job gets more difficult as the transformation ratio increases. I
seriously doubt you'll ever come close to making a transformer anything
like the one described in the last paragraph.

The amateur way is to build a system with a transformer, then figure out
how to live with whatever you get. An engineering approach usually
involves designing a system with predictable and repeatable performance,
and that precludes depending on a transformer over a wide impedance range.

Roy Lewallen, W7EL

On Oct 4, 7:59*pm, Roy Lewallen wrote:

Modeling a system like this with an ideal transformer might be an
interesting intellectual exercise. But that's all it is -- the real
system won't behave anything like the model.

Roy,

I didn't express myself well - the final paragraph was meant to be a
rhetorical question which cast doubt on the validity of the
conclusions!

73,
Steve G3TXQ

steveeh131047 wrote:
On Oct 4, 7:59 pm, Roy Lewallen wrote:

Modeling a system like this with an ideal transformer might be an
interesting intellectual exercise. But that's all it is -- the real
system won't behave anything like the model.

Roy,

I didn't express myself well - the final paragraph was meant to be a
rhetorical question which cast doubt on the validity of the
conclusions!

73,
Steve G3TXQ

Sorry, Steve. It's really hard to express subtlety or sarcasm in this
sort of written venue -- as I've found out so many times myself.

Roy Lewallen, W7EL

steveeh131047 wrote in news:5e53bd91-a69c-452f-9dab-
:

It's interesting to model this sort of arrangement, note the wide
range of feedpoint impedances that will be seen on the different
bands, and observe the feedline losses that might be incurred when it
is directly fed with 50 ohm coax. If you then introduce an ideal
transformer at the feedpoint and repeat the exercise you will
generally find that feedline losses increase on some bands and
decrease on others.

With the vertical length I tried, the effect of a 1:9 transformer was
to limit the more extreme losses at the cost of making some very low
losses higher. I guess over the several bands I tried you could say
there was a net improvement with the transformer.

But one question would be how to build this ideal 1:9 transformer
which maintains its transformation ratio and exhibits zero loss across
the wide range of impedances and frequencies involved.

Steve,

My article on the unloaded vertical includes a discussion of the unun /
balun thing. I did run models incorporating an ideal 4:1 current balun,
and found that coax loss is better on some frequencies and poorer on
others.

The model is not directly applicable to a generic end user installation
because the coax loss depends on line type, length etc, and the ideal
balun assumption is not a good estimator for practical baluns with
extreme loads.

I haven't published an article on the unun model, I should one day
perhaps.

Practical baluns are likely to have higher losses under extreme operating
conditions, and that will result in lower VSWR than otherwise, so the
added complexity of a real world balun is that at those extremes, it will
tend to be lossier, its transformation departs from ideal, and line
losses will tend to be lower.

I have created a model of a practical ferrite cored 4:1 unun, and explore
it with different core materials, dimensions, and windings. The models
reconcile well with G8JNJ's experimental ununs and my own prototypes...
but reconciliation on extreme loads taxes both Martin's and my own
measurement capabilities.

I cannot guess what inspired the application of 4:1 voltage baluns to
these antennas, much less why a reputable manufacturer would recommend
the configuration which to my mind defies sound principles. Nevertheless,
it does appear that thousands are successfully in use, and many hams have
the QSLs to prove that an antenna that lacks sound explanation "works
real good" anyway.

I do think there is good reason to apply a common mode current choke to
such antennas, not to recommend them as a matter of routine (though they
won't hurt much), but in some implementations one could expect a
significant common mode current problem, and a commom mode current choke
may be part of an effective mitigation.

As to whether a 4:1 transformation is universally better than 1:1, ideal
or otherwise, I doubt it. If the coax loss is a problem, and it will be
for all but lowest loss configurations), an ATU at the feedpoint seems
the better solution.

In making the observation that loss helps to reduce line VSWR, which in
turn reduces line VSWR, perhaps the only sensible design rationale behing
the 4:1 voltage balun applied on the subject antenna is that it drives
common mode current on the feedline, and in the case of a buried feedline
(as they often are), the power lost in heating the soil tames the
feedpoint impedance, reducing line loss. But, does that maximise system
efficiency? (Though it may work in that way, I doubt the 'designers' had
that in mind.)

Owen

steveeh131047 wrote:
But one question would be how to build this ideal 1:9 transformer
which maintains its transformation ratio and exhibits zero loss across
the wide range of impedances and frequencies involved.

Has there been any information published on loss and
transformation measurements for real world TLTs used
far outside of their design impedances?
--
73, Cecil, IEEE, OOTC, http://www.w5dxp.com

Owen,

If I've managed to read my lab notes correctly, these were the
differences in loss (ground+feedline+tuner) - with and without an
ideal 9:1 transformer at the feedpoint - for a 33ft vertical over
average ground fed with 50ft of RG213. I assumed ground losses of 20
ohms. Positive figures indicate that the losses were lower with the
transformer:

160m -1.6dB
80m +6.02dB
40m -2.3dB
30m +2.1dB
20m +4.4dB
17m +3.86dB
15m -0.55dB
12m +1.6dB
10m +2.9dB

Of course this data was for one specific scenario, but I guess you'd
look at it and say that for this case, on balance, the inclusion of
the transformer was of benefit. But now factor in some realistic
transformer losses and it might not look so clear cut.

73,
Steve G3TXQ

On Oct 5, 2:10*am, Cecil Moore wrote:

Has there been any information published on loss and
transformation measurements for real world TLTs used
far outside of their design impedances?
--
73, Cecil, IEEE, OOTC, *http://www.w5dxp.com

Cecil,

Martin has some data under the heading "33ft Verticals and 4:1 Ununs "
he
http://g8jnj.webs.com/currentprojects.htm

73,
Steve G3TXQ

steveeh131047 wrote in news:46a67bfc-c375-4533-8df0-
:

On Oct 5, 2:10*am, Cecil Moore wrote:

Has there been any information published on loss and
transformation measurements for real world TLTs used
far outside of their design impedances?
--
73, Cecil, IEEE, OOTC, *http://www.w5dxp.com

Cecil,

Martin has some data under the heading "33ft Verticals and 4:1 Ununs "
he
http://g8jnj.webs.com/currentprojects.htm

Steve,

Here are the input impedance and VSWR(50),Loss graphs for my model of a
FT240 #61 with 12 bifilar turns with a 1000+j0 load.

http://www.vk1od.net/lost/Clip045.png

http://www.vk1od.net/lost/Clip046.png

Non-ideal transformation ratio is not a big issue for an unun used with
an ATU, voltage withstand and loss are higher priority.

The balun loss data in the article at http://vk1od.net/blog/?p=568 was
obtained by measuring the balun using a VNA, and creating a spreadsheet
that solved the balun + load network for an arbitrary load impedance. The
spreadsheet is revealing, as one can immediately see the broadband
peformance of the balun with extreme loads, R and X in arbitrary
combination.

What I do know is that it is superficial to describe a balun (or unun)
with just two metrics such as 5kW, VSWR1.5... but have a look at
commercial baluns, that is how they are often (mostly) sold. There is the
odd manufacturer that gives a loss and VSWR curve on a nominal load FWIW,
but I have not yet seen any manufacturer publish a set of S parameters
covering the operating range.

I am not naive about magnetics, they are challenging devices, but at
least in the ham radio market, it is more black magic than good sense.

BTW, if you look at the loss graph for this device with a 1000+j0 load,
and assume that it can safely dissipate perhaps 20W continuous, it is
capable of less than 1kW continuous at 30MHz, but some manufacturers
build such a transformer and rate them at 5kW or more. With a load
impedance of 4k+j0 (eg a full wave dipole), the loss is even worse, and
the continous power rating even lower.

Owen