It is now 75 years since Guanella’s article “New methods of impedance matching in radio frequency circuits” described the form and operation of what we now commonly call a “Guanella 1:1” balun, but hams being hams also use other terms like “isolator”, “common mode choke”, “current choke” and some insist is it not a balun at all.
In fairness, Guanella did not call the thing a “balun”, but if we accept a very general meaning of balun to be any device intended to facilitate or permit a different state of balance to either side of itself, this is a balun.
Guanella’s description
Above is an extract from (Guanella 1944), and contains an almost complete description of the 1:1 balun. The ideal centre tapped transformers shows are a device for separating the differential and common mode currents so that appropriate elements can be used for those currents.
An LTSPICE example
Let us consider a coil as in Fig 1a with 10 turns of two wires with geometric mean diameter (GMD) of 1.6mm and coil diameter and length of 50mm and 100mm. Note that the spacing between turns is much greater than the spacing between wires, so treating it as a transmission line (TL) wound around a non magnetic tube is valid. The TL length is approximately 3m or 13.1ns, and its Zo is around 100Ω.
Above is the LTSPICE schematic. The balun is modelled with as asymmetric load (R2 and R3).
Above are the model results.
Note the upper graph of I(R2) and I(R3). They are almost equal in magnitude and opposite in phase around 41MHz, but hardly so at low frequencies.
Above are the cursor values in the previous graphic. I(R3)-I(R2) is the common mode component of current.
At 41MHz, the differential current is around 10mA, and common mode current of 1.5µA or 0.015% is very low. This is a very effective common mode choke around 41MHz.
At 30MHz, the differential current is around 10mA, and common mode current of 314µA or 3.4% is not very low and becomes much worse as frequency is reduced. This is not a very effective common mode choke below 41MHz.
Returning the the plots, the green curve is input impedance magnitude and phase. The input impedance is hardly equal to 50+j0Ω (the sum of R2 and R3) over the frequency range, and in fact varies cyclically with frequency in the upper frequency range.
At low frequencies, the impedance of the choke disturbs in the impedance transformation. In this case, the impedance below about 30MHz is too low for close to ideal impedance transformation.
The TL contributes to impedance transformation, more so as electrical length increases. When TL length exceeds 15°, it degrades nominal impedance transformation significantly, and that happens at 3MHz in this example and is 141° at 30MHz.
In summary, the example fails as a 1:1 current choke at low frequencies because the choke B is too small, and it fails at high frequencies because the TL length is to large.
Guanella did not really deal with the issue of standing waves on the transmission line, his discussion assumed matched line or intentional transformation.
How to increase the impedance of B and at the same time reduce the TL length?
The solution offered without proof is easy, introduce some higher permeability magnetic core to the choke giving higher choke impedance and with reduced TL length.
Conclusions
A good broadband Guanella 1:1 current balun has:
- high common mode impedance (typically >1kΩ for use in an antenna system); and
- near 1:1 impedance transformation over a broad range of load impedance.
Achieving both is a challenge as measures to improve one tend to degrade the other.
Air cored Guanella 1:1 baluns fall short of these objectives.
References
Guanella, G. Sep 1944. New methods of impedance matching in radio frequency circuits. The Brown Boveri Review.