Richard Clark wrote in
:

....
MFJ is not making a killing on this particular poor winding (OK, call
it historical inertia) practice, and their market for these internal
BalUns is dipoles, not monopoles.

Hi Richard,

I see MFJ getting mention, and whilst they may deserve a bit of a
flogging for some things, I spring to their defence on this occasion.

MFJ make the claim "More hams use MFJ-949s than any other antenna tuner
in the world!" and gauging from questions in online fora, they are
indeed popular, the claim may be correct.

The MFJ949E uses a Ruthroff voltage balun.

I have measured the balun losses in my '949E, and they are as I
discussed in the general case earlier, quite high on high impedance
loads. Anyone with a rudimentary knowledge of balun design would
understand that, but there are those who apparently live in an ideal who
would refute it.

The simple explanation is that most loss in practical ferrite cored
baluns usually results from losses in the core material. The losses in
the core material are related to flux density, which in a Ruthroff
voltage balun is roughly proportion to the differential voltage, and in
a Guanella current balun is roughly proportional to common mode current.
There is no reason to think that these two different types of baluns
would have identical losses.

The MFJ949E could easily be reconfigured as a Guanella 1:1 balun by
changing its end connections, but that does not make it an optimally
designed current balun. I haven't done it, and so cannot comment
further. I cannot see how the cost of manufacturing it wired as a
current balun would be any different.

I think that it is the buyers who determine the market, and savvy
sellers cater to the buyer's wants.

While anecdotal evidence abounds that 4:1 voltage baluns match up
extreme loads better, and users mostly arent't interested in finding the
root cause of the problem and fixing it, voltage baluns will be seen by
most buyers and savvy sellers as the solution.

A good demonstration of the credibility of anecodotal evidence is the
massive online support for the Array Solutions 4:1 Ruthroff voltage
balun / ZeroFive unloaded vertical combination. The configuration drives
high common mode on the coax feed line. (I should note that the sellers
recommendation has recently changed to an unun.)

MFJ is no doubt one of the savvy sellers. They do BTW have some higher
end ATUs with 1:1 current balun, as does the Ameritron label for those
who want a current balun.

My own view is that achievement of highest choking impedance in a
current balun is assisted by minimising stray capacitance to 'ground',
so I would prefer to put a current balun in a non-conductive box,
outside the ATU, on a foot of coax to the ATU. (This is one reason why I
haven't tried converting my '949E to current balun connection, the
windings are quite close to grounded metal and I expect stray
capacitance to ground is higher than desirable.)

Owen

On Thu, 29 Sep 2011 21:06:40 +0000 (UTC), Owen Duffy
wrote:

My own view is that achievement of highest choking impedance in a
current balun is assisted by minimising stray capacitance to 'ground',
so I would prefer to put a current balun in a non-conductive box,
outside the ATU, on a foot of coax to the ATU. (This is one reason why I
haven't tried converting my '949E to current balun connection, the
windings are quite close to grounded metal and I expect stray
capacitance to ground is higher than desirable.)

Hi Owen,

The internal modification would also demand a balanced tuner topology,
and not the garden variety pi configuration. Your stating stray
capacitance, coupling to ground, etc. is a preface to this.

Put simply, and as described, the 949E and its ilk are lumped
transmission line UnUn transformers. You might salvage the box in a
re-design, but it would require gutting the entire interior circuitry.

The 974 attempts to do this:
http://www.mfjenterprises.com/pdffiles/MFJ-974HB.pdf
and by their stated intentions, they are aware of the design issues.

73's
Richard Clark, KB7QHC

Richard Clark wrote in
:

....
The 974 attempts to do this:
http://www.mfjenterprises.com/pdffiles/MFJ-974HB.pdf
and by their stated intentions, they are aware of the design issues.

I am sceptical of the "true balanced tuner" approach.

For one thing, I do not recall having ever seen measurement of the
common mode impedance reported.

If the objective is current balance, high common mode impedance is
essential.

Some folk seem to think that symmetric design is the sure path to
success, but it isn't. If you take a 1:1 Guanella with extremely high
choking impedance, the currents in its output wires will be almost
perfectly balanced, irrespective of the voltage from each terminal to
ground. If you placed 10pF of capacitance from each terminal to ground,
you appear to have preserved symmetry, but the currents in those
capacitors will not be equal unless the load is symmetric. In cases
where the currents in the balun wires are almost equal and the currents
in the capacitances I mentioned are not equal, then those capacitances
have probably compromised common mode impedance.

Most implementations of a "true balanced tuner" have large stray
capacitance from each side to chassis.

But, savvy sellers will offer them to the people who are attracted by
the concept of a "true balanced tuner".

The joke of balun offerings is those designs purported to work well on
isolated loads. If the concept of the load is the overly simplistic two
terminal network, and it is isolated from ground, clearly current into
one terminal MUST equal current out of the other terminal, no balun is
required.

Owen