What is the purpose of a shunt element in a balun?

I'm working on a balun for a differential front-end receiver and see that they quite often use a shunt element (inductor) across the differential pair.

From what I could gather is that when you insert a series element into a balanced circuit, it is no longer symmetrical. Inserting a shunt element helps maintain symmetry.

Could someone please elaborate on the use of the shunt inductor (L1) in the attached balun circuit?

How do I choose, determine or calculate the value of the shunt element?

If I were to use this same balun for matching to a dipole antenna, would I still need a shunt element or will the antenna act as the shunt element?

• What's attached to the port? Does it have a capacitance spec? You may find your answer there. Commented Apr 2 at 19:18
• @TimWilliams The question stated a differential front-end receiver (CMT2310A). So the RXP and RXN pins of the receiver section. I do not have access to the impedance characteristics of the pins to determine if the pins are capacitive or inductive. Commented Apr 3 at 7:12

This circuit and the question regularly re-appears in electronics and ham radio forums, and I believe the circuit is a misused or mis-represented lattice-type LC balun, see a correct presentation in rfwireless-world's Balun basics and Balun types article

Let us redraw your circuit to match the lumped LC balun:

and simulate this balun, first without an L1 inductor:

You see a near-constant 1:0.13 graph from DC to 10MHz.

Strictly speaking, low-frequency filtering is not a job of balun, and to do this job, coupling capacitors are often used. In the design that you discuss in your question, the 330nH inductor is used to isolate low frequencies from unbalanced port:

As for recommendations on how to choose, determine or calculate the value of the shunt element, you have to learn about the antenna theory. You can start with https://antenna-theory.com/antennas/dipole.php . In short, the antenna impedance question is a bit involved. For a true half-wave dipole antenna, a feedpoint impedance consists of 73 Ohm resistance and +43 Ohm reactance (Wikipedia). The reactance frequency graph is quite steep and depends on the antenna conductor thickness; a small deviation from the exact half wavelength cancels the 43 Ohm inductive reactance. In practice, you do not calculate the parameters but use online calculators or better re-use well-established designs. The working receiver antenna can be designed with even introductory knowledge of antenna theory; the matching network for a transmitting antenna is best provided by commercial antenna tuners.

• This seems to be showing it's the opposite of a balun, i.e. it has significant common-mode gain. Notice both bal nodes are driven in phase, not oppositely. Could you discuss where the 16.5Ω tee and in-phase connection come from; is this a standard method, what justification for it, that sort of thing? Commented Apr 7 at 20:36
• @TimWilliams You're right, the narrative is untidy. In fact, what I mean to show is that, while the lumped LC balun does its job proper when it is DM-connected to an antenna (and 330nH L1 is not a random value, try and change the value to 33nH or 3300nH and you'll see), the situation drastically changes when an CM signal is fed to the balun input. It is what the L1 is added for: to fight low-frequency common-mode interference. What a star-connected 16.5 Ohm resistors do in this circuit, is a homework for you. Try and discover for yourself their purpose: a useful exercise. Commented Apr 8 at 7:28