Typical half-wave end-fed antenna kits include either a 9:1 or 49:1 toroidal balun or unun, which is a transformer for impedance matching, but with one primary winding endpoint (for the coax shield) shorted to an end of the secondary winding (the counterpose attachment) of the unun transformer windings.

What happens if this short is eliminated, making the unun into an isolation transformer? (9:1 or 49:1 or another suitable ratio to match the feedpoint impedance, perhaps with an added multi-MOhm static bleed resistor, if conditions require that).

Will antenna performance change? And if so, how?

  • 3
    $\begingroup$ There are several different ways to wire a balun, it would help if your question included a schematic rather than just a description. $\endgroup$ – user10489 Jun 14 at 4:23
  • 1
    $\begingroup$ > antenna fed with a 9:1 or 49:1 galvanically isolated transformer What do you connect the other leg of the "true isolated transformer" to? What if it were a coin cell sized battery powered transmitter, would it work without a counterpoise? Counterpoise / coax shield connection is critical, even if the connection is made through stray capacitance etc. Please draw the evolution of the transformer from 9:1 Unun to "galvanically isolated transformer" to be clear what you're asking. $\endgroup$ – tomnexus Jun 14 at 13:46
  • $\begingroup$ One common recommendation seems to be for somewhere around 5% of the antenna length in wire to be attached to the counterpose bolt on the secondary winding of the end-fed matching transformer/balun to lower the impedance to better match the impedance ratio of the transformer. $\endgroup$ – hotpaw2 Jun 14 at 15:36
  • $\begingroup$ There also seems to be a blog post somewhere regarding feeding an end-fed directly from a tiny Arduino sized transmitter, balun, but no feedline (or an inch of). Maybe LiPo instead of a coin cell was used? $\endgroup$ – hotpaw2 Jun 14 at 15:49
  • $\begingroup$ As others have said, this needs to have a schematic. As it is, it's unclear what you are asking. See my comment below. $\endgroup$ – Mike Waters Jun 16 at 13:44

Interesting. Assume half-wave resonant antenna. The feed point in the middle (just a series source, or a transformer) results in 70 Ohm real impedance. When the antenna height is lower a match to 50 Ohm is sometimes possible.

When moving the source insertion point in the direction of one of the ends the impedance grows up to beyond 1500 Ohm (depending on the environment) with a complex part in series: the capacitance of the shorter part of the line.

The asymptotic case that the "rest of the antenna length" is short as, just an example, 1 meter results in 10 pF in the chain: series impedance of counterpoise (10 pF with very low radiation resistance), feed point resistance (up to 1500 Ohm) and some losses.

In my working life I wondered about the counterpoise of a short-wave portable with a telescopic antenna of about 1 meter. The counterpoise was just the radio (without additional wires functioning as ground). In fact this set-up is a vertical dipole with very a-symmetric feedpoint, just as in your question but in total not as long as a half-wave. The counterpoise with dimensions 20 x 10 x 4 cm (portable radio) is for shortwave close to 30 pF with almost zero Ohm real. The telescopic antenna is 10 pF with very low radiation resistance: for reception it is a capacitive sensor in the E-field. Total source impedance for the low-noise (FET or tuned circuit) amplifier is close to 7.5 pF and almost zero Ohm real. (This antenna set-up is NOT for transmitters!)

So you need a counterpoise.

  • $\begingroup$ Nice answer, and I took the liberty of moving your last sentence to your profile. Is that okay with you? Somewhere here, that is sort of "frowned on" in questions and answers. (You could also change "F. Sessink" to "F. Sessink - PA0FSB". Just a suggestion. :-) $\endgroup$ – Mike Waters Jun 18 at 17:11

The coax connection is technically not a short, as the path still has to go through the transformer. Also, a 49:1 balun has a different number of turns for primary and secondary windings, where an isolation transformer has the same number of turns and is similar to a 1:1 balun. While an isolation transformer may reduce common mode current, it will not correct the impedance mismatch the 9:1 or 49:1 balun is designed to fix, and it may not have the power carrying capacity of a balun designed for transmission.

Additionally, some balun designs use two or three wires wound around the core together, with the coax shield connected to one of the wires on both ends. Removing one of these connections would effectively remove one of the windings of the transformer. Removing what you incorrectly see as a "short" does not turn it into an isolation transformer.

You can run an end fed antenna without an impedance transformer, but the result is a large impedance mismatch that must be dealt with. See What is the impedance of an end-fed half-wave antenna?

A common half wave end fed antenna is the J-pole or the end fed zepp. This antenna uses a quarter wave open wire transmission line as an impedance transformer. The transmission line (of any impedance) is shorted at one end (Z=0) and open at the other (Z=infinite), and a feed point is placed in between where Z=~50. The radiating element is connected to one side of the open end. The J-pole typically has a separate 1:1 balun, but this isn't strictly necessary. (An effective "air core" balun for this would be 5-10 turns of coax bound in a coil.)

  • $\begingroup$ See edits to the question. It’s not about 1:1 transformers. $\endgroup$ – hotpaw2 Jun 13 at 13:49
  • 1
    $\begingroup$ You still say "isolation transformer" which is a 1:1 transformer. $\endgroup$ – user10489 Jun 13 at 13:52
  • 1
    $\begingroup$ The short answer is that removing the "short" doesn't make it an isolation transformer, it removes a winding, which fundamentally changes everything. $\endgroup$ – user10489 Jun 13 at 14:33
  • 3
    $\begingroup$ @hotpaw2 Perhaps you should edit your question to include a schematic to better communicate exactly what you mean. $\endgroup$ – Phil Frost - W8II Jun 13 at 18:26
  • 1
    $\begingroup$ @hotpaw2 Shouldn't that slide be inserted into your question? I can do that for you. $\endgroup$ – Mike Waters Jun 16 at 13:37

Not the answer you're looking for? Browse other questions tagged or ask your own question.