Cebik's crossed Moxon antennas for satellite

I built a pair of Cebik's crossed Moxon antennas as described in the August 2001 QST. On the first go to put them up I miss understood the 25 ohm to 50 ohm matching section and used just a quarter wave of RG-59. The phasing line is a 1/4 wave section of RG-58 for both. They didn't work. When I put my nanoVNA on the feed line from the shack, the picture was ugly.

I went back and reread the article seeing that I should have built the matching sections with two parallel pieces of RG-59. So I rebuilt the matching sections using parallel sections of RG-59. Now I get a match of some degree on both antennas.

The problem now was that both antennas are indicating significantly low resonance frequencies.

The 70 cm antenna shows resonance at about 407 MHz. I rebuilt the 70 cm antenna believing that my impedance matching section was incorrect because the length from where I split the shield and center conductor were too long and adding length to the antenna. So I shortened them about 12 mm on each side which is about the difference between the calculated length of a 407 MHz Moxon driven element and a 435 MHz one. My resonance frequency went up only about 3 or 4 MHz. I then clipped 5 mm off the ends of all the driven elements and gained only a couple of more MHz of resonance.

The 2 meter antenna indicates resonance at about 137 MHz. I built it out of 1/4 inch aluminum rod instead of 3/16 inch aluminum rod. Moxon calculators showed that 1/16 inch greater diameter would not make a big difference in the antenna. I did the same change to the matching section. I have not messed with the element lengths of the 2 meter antenna. The resonance frequency was checked after making the matching line fix.

I know the arguments about eggbeater type antennas versus directional antennas. I looked up the standard egg beaters and noticed that they used a 93 ohm phasing section and direct 50 coax connection to one loop. I may try this as I have some RG-62 for this purpose.

Another problem is that I can find no good description of how to build phasing lines and matching sections that are not using connectors.

I don't know what I have done wrong to get to this point. Any help would be appreciated to fix this problem with these antennas. Please don't go off to using directional antennas. I currently do not have any rotaters. When I can afford an Az-El rotater, I will change my satellite system over.

73,

Bill, WD5HHH

 

A Moxon is far from being omnidirectional; it is a two element beam with around 5.5dbi forward gain and a high F/B. Crossed Moxons with phasing are likely circularly polarized and sound like a good, fairly simple design for satellite work without spending a fortune, too.

If we could see a link to that QST article we might be able to help you better. Perhaps punching in some higher frequencies into the MoxGen might get you closer to intended resonance? Just a suggestion and I'll stay tuned to this thread in case I can think of anything else.

Since the antenna will be directional I would arrange for at least Armstrong rotation for starters, too.

73,

Jeff

 

http://p1k.arrl.org/pubs_archive/101377 if you are a League member and can access it.

 

Thanks Alan. I just got back to this because of dinner and guests.

Jeff, the dimension are very close to what a Moxon generator program creates. I do understand the operation of the Moxon. Cebik modeled the performance and it was slightly better than an eggbeater or a turnstile. I just did a calculation and both antennas are about the same percentage low in frequency. I maybe need to check the frequency of the nanoVNA. I wouldn't think that the nanoVNA is over 6% low on frequency but anything is possible and I need to eliminate test equipment issues.

73,

Bill, WD5HHH

 

I finally got to check my nanoVNA and it is only about 0.01% off frequency. That is typical of the lesser grades of crystal oscillators. Is it my phasing harness causing the problem? I have seen some other designs for the coax assembly to feed crossed yagis that don't involve paralleled sections of coax. I am at a loss as to why the antennas are so far off the design frequency.

73,
Bill, WD5HHH

 

Maybe you have tried this already, but what I would do is to this:

Work on the two meter one first as it is more forgiving of small measurement errors in length and spacing.
Temporarily take off the phasing lines and the matching section and feed just one antenna element at a time directly with 50 Ohm coax and measure it with your VNA or antenna analyzer. A single Moxon should be close to 50 Ohms at the design frequency so verify that.
Record the VSWR vs frequency sweep and post a picture of it here.
Also take a picture of the antenna and feedpoint details and put them here as well.

It is a bit more difficult to measure the bigger diameter tubing accurately than the original 3/16" rod. It will also shift the antenna frequency just a bit lower than you want.
So you will need to factor that in once you measure the single antenna.

You can also measure the velocity factor of your phase and matching lines with your VNA later if you still think it is the problem.

But I think if you break the problem down into pieces you can figure out what is going on.
Good luck with your antenna !
-Harry, WB3BEL

 

Harry,

That was my next step. Only problem is I am about out of UHF connectors since I am using them on the end of my matching/phasing harness.

One thing I have found is that the supposed 70 cm antenna exhibits an SWR of 1.4 to 1 at twice the frequency near 70 cm (816 MHz versus 408 MHz). The 2 meter antenna shows a good match at about 294 MHz. This baffles me.

Pictures to be posted in a little while.

73,

Bill, WD5HHH

 

Pictures:

2 Meter :




70 cm:



I threaded the #12 AWG wire with a 2-56 thread and used 2-56 hardware.

 

. Sorry you are having problems with your antennas.However,they are great looking antennas.Good luck with trouble shooting problems.

 

I think that part of the problem is one that you identified and tried to improve already.
The length of the coax wires where you stripped them apart are too long.
They become part of the driven element making it resonate lower in frequency.
At two meters a half wave dipole will be about 7 MHz lower in frequency for each inch you add to each leg.

I think what you want is the coax center conductor and braid to go as short as possible from the split to each element.
Obviously, they should not touch the crossing element and this makes it a bit harder to do.

But I think you can make them shorter than they are currently and route them in a more direct line.
This is might be easier to do if you remove more of the center insulation from coax as it goes toward the driven element.
Just leave enough to keep the center conductor from shorting to the braid, maybe a quarter inch at most.
Or if you think you can route it shorter distance with the insulation on, then give it a try.

After you do the test with a single Moxon, and you are ready to combine the pair, double or triple check that your wiring for the phasing lines and matching section match the diagram in figure 1 of the QST article.

I can't tell for sure, but it looks like the coax that you are using has foam dielectric. This will not have the same dielectric constant or velocity factor as the original article.
So your phasing lines and matching line will need to be longer than the ones in the article if this is the case.

As far as UHF connectors go, you can reuse them if you are careful. Especially if you are just prototyping things. You can clean everything up with new if you choose when you understand what is going on.
You have to be careful not to melt the insulator in the connector if it is not the best quality. And you need to be on the look out for shorts between the coax center and braid.

Good luck with your antenna project.
Harry WB3BEL