The original transformer above comprised a 32t of 0.65mm enamelled copper winding on a FT240-43 ferrite core, tapped at 4t to be used as an autotransformer to step down a load impedance of around 3300Ω to around 50Ω.

The FT114 core has a quite low ΣA/l value (0.000505), essentially a poor magnetic geometry.

A better choice for his enclosure is the locally available LO1238 core from Jaycar (2 for $5) with ΣA/l=0.0009756/m which is comparable with the FT240 form (though smaller in size) and nearly double that of the FT114. The LO1238 is a toroid of size 35x21x13 mm, and medium µ (L15 material).

A more detailed analysis of a 3t primary winding of the effects of magnetising impedance on InsertionVSWR and system loss when it is in shunt with a 50Ω load was performed.

Above is the expected core loss.

Above is the expected InsertionVSWR.

These both look encouraging, and the next step would be to build and measure some prototypes.

Above, VK4MQ’s prototype in development. (I do not recommend the pink tape.)

]]>The original transformer above comprised a 32t of 0.65mm enamelled copper winding on a FT240-43 ferrite core, tapped at 4t to be used as an autotransformer to step down a load impedance of around 3300Ω to around 50Ω.

A very rough approximation would be that with two stacked cores, the number of turns would be around the inverse of square root of two, so 70% of the original.

A more detailed analysis of the effects of magnetising impedance on InsertionVSWR and system loss when it is in shunt with a 50Ω load was performed.

Above is the expected core loss.

Above is the expected InsertionVSWR.

These both look encouraging, and the next step would be to build and measure some prototypes.

To the original question, would half the turns be enough? No. Notwithstanding that, you are likely to find such being used, being sold.

]]>Note that the measurements are of a particular implementation and should not be taken to imply generally to 5/8λ verticals, but the solution method can be applied more generally. Lets assume that the measurement is not affected by common mode current.

The answer to the last question first is that a series inductor will not bring the VSWR much below 3. It is a common belief that a 5/8λ vertical can be matched simply with a series inductor.

There are many ways to match the measured antenna, and there are articles on this site describing some of them, but a simple and effective method in this case is the single stub tuner.

Above is a graphical solution using Simsmith. The section of line nearest the measurement load is -ve length, it is to back out the effect of the line section into which measurements were made (antenna feed point is at the cursor, 139-j191Ω). The next line section is the series section, followed by the S/C stub. In this case the series section and stub use RG213 to reduce loss. Total matching system loss is a little under 0.3dB, and the stub can easily be weatherproofed with hot glue and heat shrink tube.

One could use RG58, an exercise for the reader is to assess the loss of that option.

Obviously the length of the measurement section plays into the solution, and using its length to the mm in the model gives a more accurate result.

]]>As pointed out in the articles, the solutions cannot be simply extended to real antenna scenarios. Nevertheless, it might provoke thinking about the performance of some types of so-called balanced ATUs, indeed the naive nonsense of an “inherently balanced ATU”.

(Witt 2003) goes to some length to calculate his IMB figure of merit based on a similar load of two not necessarily equal series resistors with the mid point grounded to the ATU chassis. Witt’s IMB is equivalent to the factor |2Ic/Id|that was calculated in earlier articles in this series, and equally useless in inferring behavior in a real antenna system.

(Duffy 2010) gives an explanation of the behavior of baluns in an antenna system, and it becomes apparent that simple linear circuit solutions of a couple of resistors does not give insight into the behavior in real antenna systems.

The bottom line though is that while NEC models might be informing, there is no substitute for direct measurement of common mode current (Duffy 2011)… and it is so easy.

- Duffy, O. Dec 2010. Baluns in antenna systems. https://owenduffy.net/balun/concept/cm/index.htm (accessed 21/02/12).
- ———. May 2011. Measuring common mode current. https://owenduffy.net/measurement/icm/index.htm (accessed 21/02/12).
- ———. Feb 2012. Balanced ATUs and common mode current. https://owenduffy.net/balun/concept/BalancedAtu.htm (accessed 18/03/2019).
- Witt, Frank. Apr 1995. How to evaluate your antenna tuner In QST May 1995. Newington: ARRL.
- ———. May 1995b. How to evaluate your antenna tuner In QST May 1995. Newington: ARRL.
- ———. Sep 2003. Evaluation of Antenna Tuners and Baluns–An Update In QEX Sep 2003. Newington: ARRL.

This article reports the same asymmetric load using the MFJ-949E internal voltage balun.

The test circuit is an MFJ-949E T match ATU jumpered to use the internal balun and resistors of 50Ω and 100Ω connected from those terminals to provide a slightly asymmetric load.

The voltage between ground and each of the output terminals was measured with a scope, and currents calculated.

Above are the measured output voltage waveforms at 14MHz.

Lets work out the current amplitudes. Above, V1 (yellow) is 5.9divpp, V2 (cyan) is 7.2divpp. I1=V1/50=5.9*0.2/50=23.6mApp. I2=V2/100=7.2*0.2/100=14.4mApp.

Expanding the timebase allows better measurement of the phase difference.

V2 lags by a half cycle less 8.25µs, so V2 phase is -180+8.25e-9*14e6*360=-180+42=-138°.

Lets calculate the common mode and differential component of current in each load resistor. We will use Python as it handles complex numbers.

>>> i1=0.0236

>>> i2=0.0144*(math.cos(-138/180*math.pi)+1j*math.sin(-138/180*math.pi))

>>> ic=(i1+i2)/2

>>> abs(2*ic)

0.016100289594275147

>>> id=(i1-i2)/2

>>> abs(id)

0.01781446515461856

>>> abs(2*ic)/abs(id)

0.903776198417105

>>> 20*math.log(abs(2*ic)/abs(id))/math.log(10)

-0.8787820061070818

So, the differential component of current is 17.8mApp, and the total common mode current is 16.1mApp, the total common mode current is 90% of the differential current or 0.9dB less than differential current.

By any standard, this is appalling balance, and demonstrates why voltage baluns are unsuited to the application.

The fact that the “inherently balanced” topology is only 1.8dB better that this voltage balun experiment speaks volumes for the failure of the “inherently balanced” topology.

The measurements reported here are for a specific scenario (components, frequency and load), and should not be simply extrapolated to other scenarios.

The calculated imbalance if you like applies to the specific test circuit, and cannot really be extended to use of this balun in an antenna system scenario.

]]>This article reports the same equipment reversed so that the common mode choke is connected to the output of the MFJ-949E.

The test circuit is an MFJ-949E T match ATU followed by A low Insertion VSWR high Zcm Guanella 1:1 balun for HF. A banana jack adapter is connected to the balun output jack, and resistors of 50Ω and 100Ω connected from those terminals to provide a slightly asymmetric load.

The voltage between ground and each of the output terminals was measured with a scope, and currents calculated.

Above are the measured output voltage waveforms at 14MHz.

Lets work out the current amplitudes. Above, V1 (yellow) is 4.0divpp, V2 (cyan) is 8.0divpp. I1=V1/50=4.0*0.2/50=16.0mApp. I2=V2/100=8.0*0.2/100=16.0mApp.

Expanding the timebase allows better measurement of the phase difference.

V2 lags by a half cycle and 1.0µs, so V2 phase is -180-1.0e-9*14e6*360=-180-5=-185°.

Lets calculate the common mode and differential component of current in each load resistor. We will use Python as it handles complex numbers.

>>> i1=0.016

>>> i2=0.016*(math.cos(-185/180*math.pi)+1j*math.sin(-185/180*math.pi))

>>> ic=(i1+i2)/2

>>> abs(2*ic)

0.0013958203956907485

>>> id=(i1-i2)/2

>>> abs(id)

0.015984771545309726

>>> abs(2*ic)/abs(id)

0.0873218858170239

>>> 20*math.log(abs(2*ic)/abs(id))/math.log(10)

-21.177537875409207

So, the differential component of current is 16.0mApp, and the total common mode current is 1.40mApp, the total common mode current is 9% of the differential current or 21.2dB less than differential current.

Calculation of the common mode component of current involves the addition of two almost equal and almost opposite phase currents and is very sensitive to uncertainty in each of the measurements using this measurement method. This balun should achieve |2Ic/Id|>35dB in this scenario, but it would take a higher accuracy measurement system to measure it.

The fact that the “inherently balanced” topology measures 18dB worse that this experiment speaks volumes for the failure of the “inherently balanced” topology.

The measurements reported here are for the specific scenario (components, frequency and load), and should not be simply extrapolated to other scenarios.

The calculated imbalance if you like applies to the specific test circuit, and cannot really be extended to use of this balun in an antenna system scenario.

Continued at Inherently balanced ATUs – part 3 .

]]>LB Cebik in 2005 in his article “10 Frequency (sic) Asked Questions about the All-Band Doublet” wrote

In recent years, interest in antennas that require parallel transmission lines has surged, spurring the development of new inherently balanced tuners.

Open wire lines require current balance to minimise radiation and pick up, the balance objective is current balance at all points on the line.

Cebik goes on to give examples of his “inherently balanced tuners”.

Above, Cebik’s “inherently balanced tuners” all have a common mode choke at the input, and some type of adjustable network to the output terminals.

Cebik was not the originator of the idea, many others had written of the virtue of the configuration, but I cannot recall seeing meaningful measurement to support the claims.

Lets take the last circuit, and simulate it using A low Insertion VSWR high Zcm Guanella 1:1 balun for HF followed by an MFJ-949E T match ATU. The MFJ949E stands on its insulating feet on a large conductive sheet to serve as the ground, the balun in connected to the ATU input jack and the input jack of the balun is grounded to the aluminium sheet. A banana jack adapter is connected to the ATU Coax 1 output jack, and resistors of 50Ω and 100Ω connected from those terminals to provide a slightly asymmetric load.

The voltage between ground and each of the output terminals was measured with a scope, and currents calculated.

Above are the measured output voltage waveforms at 14MHz.

The meanings of currents used here is given at Differential and common mode components of current in a two wire transmission line.

Lets work out the current amplitudes. Above, V1 (yellow) is 4.8divpp, V2 (cyan) is 7.4divpp. I1=V1/50=4.8*0.2/50=19.2mApp. I2=V2/100=7.4*0.2/100=14.8mApp.

Expanding the timebase allows better measurement of the phase difference.

V2 lags by a half cycle and 7.5µs, so V2 phase is -180-7.5e-9*14e6*360=-180-38=-218°.

Lets calculate the common mode and differential component of current in each load resistor. We will use Python as it handles complex numbers.

>>> i1=0.0192

>>> i2=0.0148*(math.cos(-218/180*math.pi)+1j*math.sin(-218/180*math.pi))

>>> ic=(i1+i2)/2

>>> abs(2*ic)

0.011825300356025956

>>> id=(i1-i2)/2

>>> abs(id)

0.016089765935912277

>>> abs(2*ic)/abs(id)

0.7349578858466762

>>> 20*math.log(abs(2*ic)/abs(id))/math.log(10)

-2.674750918136747

So, the differential component of current is 16.1mApp, and the total common mode current is 11.8mApp, the total common mode current is more than a two thirds the differential current or 2.7dB less than differential current.

By any standard, this is appalling balance.

The measurements reported here are for a specific scenario (components, frequency and load), and should not be simply extrapolated to other scenarios.

The calculated imbalance if you like applies to the specific test circuit, and cannot really be extended to use of this balun in an antenna system scenario.

The problem starts with that it is near impossible to build such an ATU with perfect symmetry, meaning the distributed inductances and capacitances to ground are symmetric so that with a symmetric load, the entire system was symmetric and there was very low common mode load current.

Achieving that symmetry does not guarantee symmetric currents in an asymmetric load. Fig 4 and the associated text at Balanced ATUs and common mode current deal with this problem.

“Inherent balance” is a belief of the very naive… and snake oil salesmen who would relieve them of their money whilst selling the satisfaction that they have something rather special!

Superlatives like “true balanced tuner”, “fully balanced tuner”, “superb current balance” are bait for naive hams who do not test the claims, hams who do not measure the balance objective… common mode current.

Continued at Inherently balanced ATUs – part 2 .

]]>Above is Ruthroff’s equivalent circuit, Fig 3 from his paper (Ruthroff 1959). Focusing on the left hand circuit which explains the balun as a transmission line transformer (TLT), and taking the node 1 as the reference, the loaded source voltage appears at the bottom end of the combined 4R load, and transformed by the transmission line formed by the two wires of the winding, and inverted, at the top end of the combined 4R load.

It is the transformation on this transmission line that gives rise to loss of symmetry.

The complex ratio Vout/Vin is dependent on the complex reflection coefficient Gamma at both ends of the line and the line propagation constant gamma, all of which are frequency dependent complex quantities.

Vout/Vin=(1+GammaLoad)/(1+GammaSource)*e^(gamma*l)

In baluns, the real part is often close to unity, the phase is more significant.

TWLLC and family tools calculate this quantity in the long output, but I do not recall seeing it calculated in other tools. Values for Gamma may be shown in other tools, but again I have not seen gamma shown directly.

Astute readers will realise that a more correct balun could be made by including another TLT to supply the non-inverted output. By then, it looks like a Guanella 4:1 balun with the output centre tap grounded (so it behaves like a voltage balun) and has better balance… on symmetric loads.

- Duffy, O. 2007. A model of a practical Ruthroff 1:4 balun.
- Ruthroff C. 1959. Some broad band transformers. Proceedings of the IRE.

In the first article, the measurements at the input of around 7m of 50Ω line were adjusted to move the reference plane to the load end of the coax using the add/subtract cable feature of Antscope to de-embed the transmission line.

The second article used a FA-VA5 analyser and VNWA software to make the measurements and to some extent, de-embed the transmission line. In this case the transmission line was quite short at 370mm, and whilst the facility adjusted for propagation time, it did not adjust for attenuation though that was very small in this case and of little consequence. The FA-VA5 analyser and VNWA software combination would not suit the scenario in the first article as will be demonstrated.

This article examines the response to a 6m length of RG58 with O/C load at 30MHz.

We can see that although the phase of Gamma (phase of 0.85+j0.01) is close to zero the magnitude is 0.85 when the magnitude should be 1.00 for an O/C load.

The reason for this is that the port extension facility does not de-embed the attenuation of the line.

I can use TLLC to calculate the ratio of Vout/Vin for that scenario as 1.182e+0∠2.3°, so |Vin/Vout|=0.868 which reconciles well with the measured |Gamma|of 0.85.

In this case the attenuation is significant, and the port extension facility falls short in modelling the effect.

By contrast, Rigexpert’s Antscope use for the first article does include attenuation in the cable add/subtract facility and gives acceptable results.

Antscope’s cable add/subtract facility which includes a database of common cable parameters provides a convenient means of backing out the transmission line when measurements are made remotely, and whilst dependent on the accuracy of the line characteristics, it is a very useful facility not found universally.

]]>Above is the top view of the balun, and the test termination comprised two 100Ω 1% resistors clamped between the screw terminals, so pigtails were just 3mm in length.

Above is a view of the interior.The coax pigtails are quite short, they exist at the input and output.

Because of the pigtails inside the box, and on the termination, there is some unavoidable lead inductance.

Measurements were made with a FA-VA5 and VNWA PC software.

A measurement of impedance looking into the balun is obtained by using port extension to subtract the effect of the known length of RG316 transmission line.

To be more correct, it de-embeds the propagation time of the additional line but it does nothing to de-embed the attenuation which is quite small for the line in this case, so the error is small.

Above is the scan from 1 to 31MHz. The existence of some inductance at both ends of the transmission line complicate the results, but it appears that the port extension correction is approximately correct and Insertion VSWR reaches 1.1 at 30MHz, some of which is due to imperfection in the termination used. Nevertheless it is a good result, partly due to keeping pigtails short, partly that the screw terminals allow a short termination. Equivalent series inductance would appear to be perhaps 30nH.

Classic baluns for antenna use tend to have screw terminals that are widely spaced, often on opposite sides of the box, seemingly to suit wider spaced transmission lines. Of course when used with such lines, Insertion VSWR is not relevant, but where they are used with say, a HF Yagi, wide spacing is sub optimal.

]]>