0
$\begingroup$

I am wanting to build 2 x 4 element vertical yagi antennas for 28 MHz and co-phase them and feed with one 50 ohm coaxial transmission line.

If i were to use a gamma match for each yagi and tune them each for 50 ohms, then use 50 ohm coax of the same length from each antenna joined together so the antennas are fed in parallel, that would give 25 ohms.

Can anyone tell me how to make a 1:2 balun to match the 25 ohm 2 x yagi antennas (balanced) in parallel to the 50 ohm coax (unbalanced) ?

There is of course plenty of information around about the reverse, how to make a 2:1 balun, the opposite of what i need.

Or, i could tune the gamma matches for 100 ohms and then those in parallel would give 50 ohms, but i'm not sure if a gamma match will allow adjustment for 100 ohms, does anyone know about this ?

Any advice appreciated.

Improve this question
$\endgroup$
6
  • 1
    $\begingroup$ A little schematic of the desired architecture would be greatly appreciated! $\endgroup$
    – Wireless Learning
    Commented Mar 31, 2022 at 12:13
  • 1
    $\begingroup$ How about a 1:1:1 balun? (3 windings on one toroid) $\endgroup$
    – hotpaw2
    Commented Mar 31, 2022 at 17:31
  • $\begingroup$ The usual advice for co-phased stacks is to use $\lambda4$ lengths of 75$\Omega$ line to act as transformer sections that tee into the 50$\Omega$ feedline. I guess this interferes with gamma matching, which is why not an answer. $\endgroup$
    – user21417
    Commented Mar 31, 2022 at 20:03
  • $\begingroup$ @hotpaw2 I was thinking about that, a 1:1:1 balun, do you have a drawing of this ? also will that give 25 ohms to 50 ohms impedance transformation ? $\endgroup$
    – Andrew
    Commented Mar 31, 2022 at 23:35
  • $\begingroup$ @jvd A coaxial transformer has the disadvantage that it only works properly at one frequency and has a narrow bandwidth, and so limits the bandwidth of the entire system, the yagis i am using are designed to be wide band and cover about 2 MHZ, a toroidal transformer is somewhat frequency independent and much better, but thanks for the comment ! $\endgroup$
    – Andrew
    Commented Mar 31, 2022 at 23:37

1 Answer 1

Reset to default
2
$\begingroup$

A gamma match on the yagi should give you 50 ohms unbalanced. If you do it both sides and have a balanced input, it's usually called a T-match.

So the two Yagis each provide a roughly 50 ohm unbalanced feedpoint, which can be extended by any length of coax.

The usual solution to the matching is to use quarter-wave sections of 75 ohm coax - RG-59 or similar, on each antenna. This transforms the 50 ohm antenna to about 100 ohms, and two of these in parallel is 50. The transformer impedance should really be $Z_T=\sqrt{Z_1/Z_2}$ or 70.7 ohms but 75 is close enough and readily available.

A better idea, given construction tolerances, is to first bring the two antenna coax cables together to form a 25 ohm feedpoint, and then use a parallel pair of quarter-wave 75 ohm cables which form a 37.5 ohm coax, to bring it back to 50 ohms.

The quarter-wave transformer will work fine over a much larger bandwidth than a Yagi. You can calculate it with the online Smith chart at https://www.will-kelsey.com/smith_chart/. I started with a 100 ohm load, 28 ±5 MHz and applied a 2.68 metre 75 ohm transmission line. The SWR to 50 ohms is below 1.5:1 from 24 to 32 MHz. It is possible to get wider bandwidth

Note that if your antennas have different impedances, the power split between them may not be equal. For example if they both have an SWR of about 1.5:1, but one of them is 75 ohms by the time it reaches the parallel point, and the other is 37 ohms, then the power will split in a 2:1 ratio. So ideally you would tune the gamma match with a VNA. Of course if they're 1:1 then it will be fine.

Also there will be coupling between the antennas, unless they're very far apart, which will change the feedpoint impedance slightly, but both in the same way. Again, it's possible to measure and estimate this with a VNA.

Finally, it's best to measure the quarter-wave sections yourself, don't just cut from the published velocity factor. Another reason to have a network analyser.

Improve this answer
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .