Voltage baluns are making a comeback for HF antennas?

I planned this article to be a Youtube video, but recent behaviors of Google / Youtube give me pause to consider whether or how I use Youtube in the future. So, this article is a simpler presentation of the underlying concepts.

For most purposes, users of HF antennas would like the feed line to perform that function alone, ie to not participate as radiating or pickup conductors.

For that end, we want the common mode component of feed line current to be very very low.

Discussion of antennas tends to represent them as two terminal devices in free space, ie ignoring the presence of ground in close proximity. This applies whether the feed line is coax or two wire line.

A more complete representation (or model) is a three terminal network that includes a terminal to permit current to flow to ground. Above is a Wye or Delta equivalent circuit, and that can be transformed to an equivalent Tee circuit.

Let’s explore a Simsmith model of a practical and common 1:4 Voltage Balun (A Ruthroff 1:4 balun) with such a three terminal load and calculate some key metrics of the load ‘balance’. The model has a parameter that allows interactive tweaking of the position of the ground tap on Z1+Z2=200Ω.

Above is a screenshot of the Simsmith model. Z3=1GΩ, so essentially the load is two series resistors Z1 and Z2 with the connection between them grounded. This type of load configuration is used in some magazine articles discussing baluns and may be familiar to the reader. In this model, the sum of Z1 and Z2 is specified as zt (=200Ω) and the tap point specified as a portion of the total (tap=0.4 in this case). The calculated three terminal network is shown as z1,z2 and z3 (80, 120 and 1G) in this case.

The balun modelled is a Ruthroff 1:4 balun comprising 7t windings on a FT240-43 equivalent toroidal core, a very common balun. The magnetising impedance and transmission line effects are modelled.

Above is a chart of some key results. It is a bit busy, but lets work through each of the four plots.

The solid blue plot is InsertionVSWR. This is a 1:4 balun with a 200+j0Ω load, so it should have very low InsertionVSWR. InsertionVSWR departs from ideal due to two main factors:

  • magnetising impedance at low frequencies; and
  • transmission line transformation in the pair of conductors forming the bifilar windings.

InsertionVSWR would probably satisfy all but the most demanding users.

Let’s look beyond VSWR, and the function of this thing is to connect a nominally, but slightly poor balanced load to the source.

Voltage balance means equal magnitudes and opposite phase voltages (ie 180° phase difference) at the output terminals.

The dashed magenta curve is the ratio of the magnitudes of V2/V1 (%) (V2 and V1 are the output terminal voltages. For a good voltage balun, we would expect that this is very close to 100%. In this event, it starts of very close to 100% at 1MHz, but falls to 91% at 30MHz. The magnitudes are significantly different,  ‘voltage balance’ is well away from ideal.

The dashed red curve is the phase of V2 wrt V1. For a good voltage balun, we would expect that this is very close to 180°. In this event, it starts of very close to 180° at 1MHz, but falls to 146° at 30MHz. The phase is very different,  ‘voltage balance’ is well away from ideal.

Now the important target is current balance. Current balance means equal magnitudes and opposite phase  currents (ie 180° phase difference) flowing in the output terminals.

The dashed blue plot shows the ratio |2Ic/Id|, the magnitude of the total common mode current to the differential component of current in the output terminals. For a good ‘current balun’, |2Ic| << |Id|, perhaps |2Ic/Id| might be 0.1%, likewise for a good voltage balun with a symmetric load.

But this load is asymmetric (80+120Ω) and at 1MHz |2Ic/Id|is 40%, and rises to 69% at 30MHz.

Conclusions

This study shows that even for a modest imbalanced load on a practical voltage balun, common mode current is relatively high.

If you think your load is balanced but you have not measured it, you really don’t know.

At the end of the day, if the target is reducing common mode current in an antenna system to an acceptable level, directly measure common mode current in your installation. Note that common mode current is almost always a standing wave, so sampling at multiple points is needed to assess the magnitude, and it is almost certainly frequency dependent. This is a whole lot more meaningful than contrived brag factors like CMRR espoused by some authors.