Coax Power Dividers for Two or Four 50 ohm Antennas

 by Tom, K1JJ 


Hola Guys,

Stacking antennas and  phasing them together is not as hard as you may think. Whether they are phased dipoles or Yagis, here's a simple method of feeding for the proper power split and match to 50 ohm line. Four port = feeding four 50 ohm antennas.    Two port = feeding two 50 ohm antennas.


"Credit given to Mike, W6MYC from M2 Antennas for supplying this great information."
Tom, K1JJ
------50.1 Mhz examples:

A four port coax cable power divider is made of two pieces one-quarter wavelength long and with a 'T' in the middle and at each end. Theory: on each side, you bring two 50 Ohm lines together in parallel bringing the impedance to 25 ohms. 25 ohms passes through one-quarter wavelength of 50 Ohm cable and becomes 100 ohms at the other end. The same thing happens on the other side so at the center 'T' you have two 100 Ohm impedances coming together in parallel and you are back to 50 ohms. This system uses no 75 Ohm cable.

Your other question was how long the 50 Ohm phasing lines should be with the antennas configured in a collinear fashion. The simplest way to do it is to keep the lines in one wavelength multiples. Like the inner two Yagis might be fed with 2.5 wavelength cables and the outer Yagis would then be fed with 3.5, 4.5 or 5.5 wavelength sections of coax.

Here is your formula.

11802.8/ freq in MHz = 1 wavelength in inches in free space. So 11802.8 / 50.2 = 235.116 inches. For one-quarter wavelength you divide that number by 4. 235.116 / 4 = 58.779 inches.

When using coax you must multiply this 58.779 times the velocity factor of the coax. Let's say you are using LMR 400 for all your lines like I do. LMR 400 velocity factor is 0.85 so 58.779 x 0.85 = 49.962". When we build this 4-port divider, we put an 'N' connector on one end of about 51" of LMR 400 and using a sweeper and scope we trim the cable until it is a one-half wavelength resonant cable at 50.2 x 2 or 100.4 MHz. Now in actual practice, we trim about one inch off this final length to compensate for the length of one-half of the center 'T' connector. This system is several MHz wide to the 1.2:1 VSWR point,s so if you miss by 0.5 MHz it won't matter at all in the big picture. The most important thing is that both pieces are the same electrical length (i.e. they are resonant at the same frequency).

Use the same basic formula when cutting the phasing lines. One wavelength is 235.116" x 0.85 = 199.85 inches. Put that number in your calculator memory. After you have the minimum length of the cable required to reach from the center power divider to the outer and inner Yagi feed points, divide this 199.85 into the length you have come up with. Let's say the outer length needs to be 65 feet or 780 inches. So 780 / 199.85 = 3.902 wavelengths. This tells you to make the cable just a bit longer or 4 wavelengths 4 x 119.85 = 799.39" or about 67 feet to start before trimming to exactly four wavelengths at 50.2 MHz.

You can use an MFJ antenna analyzer to trim your cables. Put a 'T' on the MFJ. On one side you terminate with a 50 Ohm load. Connect your cable with one connector on the other side of the 'T' and 'sweep' the MFJ tuning for the best match at around 50 MHz. Now go down in frequency until the VSWR is 1.5:1 and note that frequency. Now go up in frequency until the VSWR if 1.5:1 and note that frequency. The midpoint between the two 1.5:1 VSWR frequencies is your cable's exact one-half wavelength multiple point.

At each one-half wavelength multiple an open cable reflects an open back to the 'T' on the MFJ, so the MFJ only sees the 50 Ohm load. At any other frequency other than a multiple of the first, the cable introduces every other impedance all the way up to a short which occurs on an odd one-quarter wavelength long cable. To summarize, an open  one-quarter wavelength section reflects an RF "short" back to the 'T' on the MFJ and an open one-half wavelength section reflects that same open back at the 'T'. The MFJ does not like to 'see' a short, so we go to twice the frequency where that cable is now one-half wavelength long and it now reflects an "open" at the 'T' as if it were not even there and the MFJ just 'sees' the 50 Ohm load again.

Phasing two 50 Ohm antennas together is done the same way as before, but you must use two one-quarter wavelengths of 75 Ohm cable like RG-11 on either side of the center 'T'. Barrels go at the outer end of the one-quarter wavelength sections to couple to the two equal length 50 Ohm phasing lines.

I should mention that you do not need to be in one wavelength multiples for the phasing lines in this setup; just two equal lengths. Back to the 4-port system we discussed yesterday. If you had a classic configuration, where four equal length phasing lines could be used, you would not have to make them any particular wavelength multiple. You could just cut them all equal and be done with it. We like to do things at one-half and one wavelength multiples, so we can be sure they are all "electrically " the same length. Another small advantage of using one-half and one wavelength multiples is that the true impedance of the antenna feedpoint is mirrored at each one-half wavelength down the line so when you look up each individual phasing line you see the real antenna except for some slight improvement due to feedline loss. In a good quality phasing line this effect is is minimal.