W2DU -Staggered Dipoles (Posted for Walt)

[W1UJR]

Hi All,

I discussed staggered dipoles in Chapter 20 in Reflections, but I didn’t take or publish any measurements. I’m speaking about staggered 80 and 40 meter dipoles for use with a single coax feedline on 80, 40, and 15 meters. Lately I’ve heard several people complain that when adding a 40m dipole to the same terminals as the 80m, and positioned directly below the 80m, the 80m dipole needs retriming to maintain resonance at the desired frequency. I have not found this to be necessary. Consequently, measurements appeared to be necessary.

Just this past week sons Rick, W8KHK, and Bill, W2WM and I made the necessary measurements. Bill and Rick constructed the staggered-dipole test antenna, 125’ for 80m and 66’ for 40m at roughly 27’ above dry sandy ground. The 40m dipole was strung approximately 2’ beneath the 80m dipole. I calibrated the SF-214 feedline, and Rick took all the measurements using a General Radio GR-1606A RF impedance bridge fed with an HP-8640B sig gen, and detected with an HP-8405A Vector Voltmeter. The measurements taken were to determine the differences in dipole terminal impedance and mismatch between dipole and feedline between the 80m dipole alone and with the 40m dipole attached.

The feedline calibration included attenuation and electrical length at each frequency of measurement, and characteristic impedance. Thus the final data relates to the dipole terminals, not the input to the feedline.

The resulting data and graphical plots are in Excel format, included here as two attachments. It should be noted that there was no fudging of any of the data points—all the measured data fit the curve, attesting to the accuracy of the measurements that you’d expect from the General Radio Bridge as compared to the MFJ-269.

As I said earlier, CQ Magazine is currently working on the publication of Reflections 3, and this new data will appear there as an extension of Chapter 20. Hope you find the data useful.

Walt, W2DU

[W2DU]

I was unable to post the above data, so I asked Bruce for help. It's possible that the problem was that my files were over sized. The data in the views above are too small to read easily, so I've reduced the size of the data files, with another go at posting again to see the data more clearly. Here goes.

Walt

[W2DU]

Sorry, for some reason the data failed to come through, so I'll copy it to this location

Data to accompany graphs of R + jX and SWR
            
   Z = R + jX of Stagger-Tuned Dipoles
f in MHz R 80      R 80-40      X 80     X 80-40

3.5           55.25        51.55        -89.43        -82.8 3.55           57.72        55.2      -68.84       -65.28 3.6           60.78        58.17      -48.08       -43.86 3.65           64.53      59.93      -26.94       -22.77 3.7           67.06        66.02        -4        -1.35 3.75           69.69        71.34       16.73        20.61 3.8           75.08        77.04       40.91        44.4 3.85           79.32        82.56       64.34        68.75 3.9           84.61        90.73       87.34        92.84 3.95           91.02        99.89      111.4       121.13 4           98.49      109.62      135.18       149.02

SWR of Stagger-Tuned Dipoles

f in MHzSWR 80m   SWR 80-40

3.5            4.468   4.396

3.55            3.336   3.22

3.6            2.359   2.221

3.65            1.688   1.556
3.7            1.334   1.303
3.75            1.526   1.617
3.8            2.125   2.237
3.85            2.887   3.043
3.9            3.782   3.97
3.95            4.835   5.182
4            5.949   6.466

[/table]

W2DU  10-23-09

[W2DU]

Perhaps can someone please assist in lining up the data in the post above? The alignment was perfect prior to posting.

Walt

[kb3ouk]

Data to accompany graphs of R + jX and SWR
            
   Z = R + jX of Stagger-Tuned Dipoles

f in MHz      R 80        R 80-40           X 80           X 80-40
3.5             55.25       51.55            -89.43          -82.8
3.55           57.72       55.2              -68.84          -65.28
3.6             60.78       58.17            -48.08          -43.86
3.65           64.53       59.93             -26.94          -22.77
3.7             67.06       66.02             -4               -1.35
3.75           69.69       71.34              16.73            20.61
3.8             75.08       77.04              40.91            44.4
3.85           79.32       82.56              64.34            68.75
3.9             84.61       90.73              87.34            92.84
3.95           91.02       99.89              111.4            121.13
4               98.49       109.62             135.18          149.02


SWR of Stagger-Tuned Dipoles

f in MHz       SWR 80m           SWR 80-40
3.5                  4.468              4.396
3.55                3.336               3.22
3.6                  2.359              2.221
3.65                1.688              1.556
3.7                  1.334              1.303
3.75                1.526              1.617
3.8                  2.125              2.237
3.85                2.887              3.043
3.9                  3.782              3.97
3.95                4.835              5.182
4                    5.949              6.466



Hope this is better, Walt.
Shelby KB3OUK

[kb3ouk]

After editing it about a dozen times, I think I've got it right now.

[W2DU]

Thank you, Shelby, good job!

Walt

[Pete, WA2CWA]

If you just copy a set of cells from a spreadsheet and bring them over here to paste them, it generally results in a jumble of data.
Taking some of Walt's data and just entering it manually into Excel, I first adjusted the data to have two decimal places. I then adjusted all the columns to have the same width. When posting a message reply, I used the "Insert Table Row" command "tr" and "/tr" to copy all the rows in the block of data. You can also use the "Insert Table Column" command "td" and "/td" and copy all the columns in the block of data. The "Insert Table" command never seemed to work for me. Doesn't come out perfect but it's a lot easier then adding and deleting blank spaces all over the place to make the columns sort of line up.

f in mHz   R 80      R 80-40      X80      X 80-40

3.50      55.25      51.55      -89.43      -82.80 3.55      57.72      55.20      -68.84      -65.28 3.60      60.78      58.17      -48.08      -43.86 3.65      64.53      59.93      -26.94      -22.77 3.70      67.06      66.02      -4.00      -1.35 3.75      69.69      71.34      16.73      20.61 3.80      75.08      77.04      40.91      44.40 3.85      79.32      82.56      64.34      68.75

[Steve - WB3HUZ]

Using the code or typewriter command will help too.

Good data. I never noticed any problems when using a combo of 80-40 meter dipoles with the two spaced less than six inches.

[K1JJ]

When you model a single 75M dipole (figure 8 pattern) and then add the 40M dipole, there is a very slight change in the 75M pattern. It isn't much, but evidently is also diplayed in the very slight variations in the swr readings posted here.

T

[KX5JT]

Good timing on this post.  I just ordered a 100' of RG-213 with the idea of making a "Fan dipole" with 80/40/20 meters all fed by the single coax.  (Theorhetically it should also give me 15 meters via the 40 meter portion.)

So.  With mostly AM operations in mind, should I use a 1:1 choke?  I have a w2du style unadilla but eventually when I get MO' POWAH will it be okay?  Should I forget it and wrap some turns of the coax around a pvc form for a choke?  Should I forget the choke all together and just feed the antennas directly with no contraptions at the feedline?  Should I keep the distance CLOSER than 6 inches (as implied by Steve) or should I separate the dipoles out farther?

Any advice will be considered and appreciated.

[K1JJ]

John,

I like your avatar...


Any of the techniques you described will work FB.  There is a great need for a balun(s) on a multi-element array, like a Yagi or stack where undisturbed phase is critical. There is less need on a simple dipole like you are planning.

Personally, if it were my fan dipole, I would wind about 6 turns of coax on a 4" form and also slip on some ferrite beads (maybe 8-10 beads) right at the feedpoint before the coax turns.  This will give you a high impedance on those bands.

I don't know what the power rating is (avg or pep) on the W2DU balun, but I'm sure Walt will know.

The spacing of the wires of a fan are not all that importatnt. I know guys who used zip cord for a fan dipole - or 300 ohm twinlead. Those are spaced a fraction of an inch. When considering that openwire feeders can be spaced as much as 2 feet apart on 75M and still cancel, that tells you the true story.  ie, Just about any spacing you do between wires will couple and is nothing to be concerned about.  You will be pruning the individual wires for best swr match anyway.  The wire with the lowest impedance will accept the most current. That's why the 40M dipole also works on 15M, cuz it's a low impedance feed on that band, though exhibits a cloverleaf pattern.

T

[W2DU]

Hi John,

Glad my post concides with your project. I was hoping it would be of value for someone.

You are right, the fan dipole will work just fine, with the 3rd harmonic of the 40-m dipole working well on 15m. However, the radiation pattern on 15m will be multi-lobed, a four-leaf clover pattern, the same pattern as the 102' dipole of G5RV on 20m.

Now concerning the balun. If you were taking impedance measurements the balun is absolutely necessary to eliminate common-mode current from the outside of the outer conductor of the feedline. That current destroys the accuracy of the impedance measurement.  However, if you omit the balun and have no problems with RF in the shack, or RF biting your lips when you touch the mic, then go without the balun. The common-mode current on the feedline will affect the radiation pattern somewhat, but not significantly on the lower frequencies.

I've never been thrilled with the quality of the W2DU baluns made by Unadilla--those made by the Wireman are far superior. Some people have reported overheating with the Unadillas, but Press Jones, the Wireman, has had no reports of overheating with those he made. I've forgotten the name Unadilla gave to their balun for frequencies higher than 75m, but they ignored engineering on that balun by simply reducing the number of 73-mix beads from 50 to 25. They still called it the W2DU balun, but those baluns were bound to overheat at QRO.

I don't think winding the feedline around a PVC is a good option in this case, where you're operating from 75 to 15m, because the inductance required for 75m is much higher than that required for 15m. I'm not sure that the 75m coil balun would be as effective on 15m as one constructed particularly for 15m.

I believe that RG-303 or RG-141 will handle all the power you'll use, so here's another suggestion. Wind four or five turns of either of these coaxes through a large ferrite toroid of the proper mix, such as 73, 43, or 61, and connect this coax between the load end of the RG-213 and the dipole terminals. I'm sure this arrangement would have no heating problem with QRO.

Concerning the spacing between the various dipoles, it won't make much difference whatever the spacing. Steve is correct, 6" will be fine. I've also known some people to use ribbon rotator cable for this purpose, cutting each conductor of the four-conductor cable to the correct length for each band.

Hope this helps, and let us know the results when your setup is finished  and tested.

Walt

[W2DU]

I just now recalled the name Unadilla gave to their poorly-engineered balun with the number of beads reduced from 50 to 25.  

They named it the "Beam Balun" for use on 20m and above, apparently thinking that at the higher frequencies fewer beads were required. They should have looked at the impedance vs frequency plot appearing in the graph Fig 21-3 in Reflections to discover that the series impedance of a balun with half the number of beads wouldn't be sufficiently effective in reducing the common-mode current, which would be higher than with 50 beads, and thus increasing the I^2R loss and dissipated heat.

Walt

[WA1GFZ]

Check the minimum bend radius on the RG141 size teflon coax before you wind it around a torroid. Might be better to stack a pair of them.  It will handle a lot of power.

[WQ9E]

Thanks Bruce and Walt for the interesting data.

I never had a problem using 80 and 40 meter dipoles together like this but I was less than wildly successful when I also tried to add 20 meters.  My first antenna as a novice was a dipole cut for 80 and 40 made from flat rotor cable and fed with RG-8 without a balun (I changed to this after trying to get a 40 meter dipole to work on 80 as a T).  It was attached to the second story roof running along the sloping eaves and then out to tress and it worked OK.  My Valiant had no problem loading this antenna up to the 75 watt novice limit on 15 meters also.  

My third week as a novice VQ9MI from the Seychelles responded to my CQ on 15 meters and this was my second DX contact (the first was VE land on 40 meters which was a fair distance from my Gulfport, MS QTH).  Two weeks later I worked VQ9DC, the other half of the husband and wife team.  Those were very fun times!

Rodger WQ9E

[KX5JT]

Thanks Tom and Bruce.

Given the information received here, I will first just throw the fan dipole up, tune each leg to the proper "watering holes" for the bands and feed it directly with the RG-213.  I will have a pulley system for easy access, so I will see how that plays.  If once that is done, there is RF in the shack, then I'll decided on a choke/balun scenario.  I'm a firm believer in keeping it simple.

[W2PHL]

Staggered Vertical?

I use a 33' elevated vertical with four downward sloping radials for 40m. As you might expect the feed-point impedance is sky high on 20m. On a hunch, I taped a 17' piece of wire to the outside of the 33' fiberglass pole that houses 33' of wire(vertical element on 40m). I also added four 17' radials. I had to cut the 17' vertical element down to about 15' but I got the feed-point down to about 75 ohms, swr about 1.9:1 at 14.2MHz. So just to be clear, the 15' and 33' vertical elements are tied together at the feed point and spaced about an inch or two (it varies) apart. The elevated ground radials (four 33' and four 17') are also tied together. I didn't notice any significant changes on 40m on the mfj-269 after adding the additional elements.

I assume there must be some coupling between the vertical elements which would account for the shorter than expected (15') vertical element on 20m. I wonder what effect the 33' vertical section is having on 20m. Is it possible the 33' wire is acting as a 1/2 wave on 20m and the 15' section is not acting as a 1/4 wave vertical but more like a method of coupling for the 33' vertical?

[K1JJ]

Staggered Vertical?

 
I assume there must be some coupling between the vertical elements which would account for the shorter than expected (15') vertical element on 20m. I wonder what effect the 33' vertical section is having on 20m. Is it possible the 33' wire is acting as a 1/2 wave on 20m and the 15' section is not acting as a 1/4 wave vertical but more like a method of coupling for the 33' vertical?

Good idea.


The section that presents the lowest impedance will "hog" the current at the feedpoint.  It will divide approximately based on the impedance ratio.  So yes, some current will flow in the 40M vert on 20M, but since 40M is a 1/2 wave high impedance feed on 20M, it will pull little current.  There is always SOME pattern  distortion away from a 1/4 wave perfect model due to this current, but usually insignificant.  

Probably one of the best innovative verticals is the SteppIR vertical which feeds a metal tape up and down inside a fiberglass hollw tube for tuning.  That's like putting on a whip tuned for each band.

T

[W2PHL]

That SteppIR vertical is neat. Too bad the roof is the only place I would have to mount it and elevated verticals like tuned radials so that would quickly get out of hand.

Back to the staggered dipoles.... I'm going to try a 160/75 combo soon. Half of the 160 legs will be zigzagged around the yard. I always wanted to give 160m a try. I hope the extra wire doesn't raise the QRN on 75m.

Phil