The AM Forum

See the attached file.


The actual antenna I built was a little different.

For one, I did not need the series matching section after the switch box as shown in the drawing. The impedance was already around 45 Ohms at the switch box, so adding the series section would have only made it lower.

Second, I had to trim several feet off each loop. Originally, the system resonated at around 6.9 MHz. I moved it up to around 7130 kHz and stopped trimming. I don't know if this was some short coming and/or mistake in my modeling or because I used insulated wire. Anyway, the dimensions shown in the drawing are probably a good starting point.

Finally, I used a slightly longer stub (actually two - one for each element) on the reflector element. I remodeled my system at a slightly lower height, about 65 feet at the top versus 70 in the original document. With this setup, I achieved a better front-to-back F/R ratio with about 140 Ohms inductive reactance instead of the 120 Ohms in the original design. So I made the stubs a little longer. If one could make some field measurements on the F/B, you could trim the stub for best performance. I'm not going to worry about a few dB at this point. You can see the pattern with the optimized stubs in the attached image file.

Steve...FB OM!

Thank you very much for the info on the loops!  We have previously danced all around the details in our discussions.   You are very considerate to share your findings and the details, and it is certainly appreciated.  Your antenna really works well and we can all learn from your experience.

My loops presently are diamond-shaped and one is about 5% shorter.  I am presently using that one as a director.  I plan to try switching a stub or additional length on that element to change from director to reflector.  The loops are mounted on a short cross arm with pulleys on the tower with the apex being at about 74 ft. AGL.  The spacing is only about 4 ft. at the top, increasing to approx 20-25 ft. at the bottom of the loops.  This is described as a "pyramid" or "triangle" antenna in some books, except my elements are diamond-shaped instead of triangles.  I am feeding the lower corner of the driven element with RG-11 in a matching section, then 9913 to the end of the 1/2" hardline at the base of the tower.  Then about 240 ft total run of 1/2" 50 ohm hardline to shack. .  I have tried feeding both loops thru a harness/matching section with a tee, but the dimensions were not optimized.  My SWR went up from 1.2 or so up to about 1.7 when I went to the combined feed.  That was just an experiment to see if I noticed any great difference, which I did not.

I do not yet have the transfer relay box installed, nor have I optimized the line-up and trimming of the loops themselves.  However, the bottom-line is that they really seem to work!  When I get the switching going, I will have a much better picture of just how well they work, but reports into Europe and other DX have been quite good, certainly better than a plain dipole.  I also plan to install a 20 ft boom on the tower and then maintain a 20 ft spacing overall between the loops.

Anyone who wants to really enjoy working DX on the low bands should try one of these antennas.  They work very well and allow you to be quite competitive compared to big yagis, etc.

Thanks again Steve!

73,  Jack, W9GT

I

Yep, after witnessing the Huzman work Indonesia on 40M the other night, I KNOW his radio talks, caw mawn! 
The W4's are still scratching their heads over that one.

(Can I get a witness? Can I an amen?   ;D)


T

Yup...I wuz there  8) ;D  And Tom was trying to look-up where YC6 is at.  Hi Hi FB OM.

73,  Jack, W9GT

How critical is the distance between loops? 20 ft is good , but is 30 ft better for forward gain? Have you done any modeling on the spacing? ..I would like to try a set supported by pine trees....single point at top delta loop  bottom feed ..I can get maybe 70 ft at the apex..Great job Steve.... lots of good info....I like hearing about the 40 meter DX...

Greater spacing does not increase gain. It will likely reduce it for a parasitic array. If you make it a phased array, you could just change the phasing to make it work.

If you can get it up at 70 feet, you will kick a$$!

Steve.  You said you had to trim the length "several feet" to get the loop to resonate...approx how much?  Also could you give some specs on the caps and RFC and K1 in the switch box? What kind of relay did you use for the switch box?? I want to feed my loops at the bottom...If the triangle has the pointy end up with all legs approx equal, and I feed in the center of the bottom horizontal leg, I think I should get horizontal polarization..

I'm feeding center of bottom and point up. My loops are about 140 feet lotal length each. I have a quarter wave of RG-11 (solid dielectric VF .66) 20.5 feet hanging off each feed point.
I ended up switching a 9 foot shorted stub to the undriven element. This works well but now I have a director rather than a reflector. This is because the stub is after the quarter wave section. I tried an open stub but resonant frequency shifted a lot. I want my switching close to the ground so will need to add another 20.5 feet to the stub to flip it back to a director.
I have not tried force feed yet but with the two feeders within reach anything is possible. The bottom flat side is about 60 feet long with the apex at about 55 feet.
I've had reports and have seen 3 s units FB. A reflector will improve FB.
I would like to add a third driven element in the middle and stub the two outside elements but have not made simulation work quite right yet. 

GFZ...How far apart are your loops?  I am currently surveying the 40 or so 80-100ft tall pine trees that surround my shack for the best location for this antenna....Of course the trees closest are already festooned with antennas...so I will wind up with some feedline going into the shack...I hope to be able to get the apex up around 70 ft... I would be interested in your force feed results....Good Luck...Steve

Yes, you will get horizontal polarization.

I don't remember the exact amount but it was several feet - more than I expected. However, the trimming brought the resonant frequency up and I was able to get it where I wanted it to be. Other than some time consumed, it wasn't hard.

I did not use the exact switch box design shown in the design document. Instead I used separate DC control lines. The one switch box shown allows you to run the DC control voltage over the coax. So, you would need a cap that would handle the peak voltages found at around 50 Ohms and at the power you plan to run (the cap should also be able to handle the current too). Add in a 2-10x safety factor and you are set. A typical 2.5 uH type used on the output of most pi-nets should be good for the choke.

I used two double-pole, double-throw relays for the switch. The relays were about 2-3 inches tall by an inch wide and an inch deep. Most any relay with good sized contacts (IIRC mine were 20A) should be just FB. I used ones with a 120 VAC coil and ran them on 24 VDC.

Switch box photo.

My loops are also 20 feet apart hanging from a 20 foot boom on the tower.

Steve, Do you think the common shield at the switch box effects anything?
I was considering floating the connectors on some G10 so the shields of each antenna only got terminated to the box from the feed side. This way the stub was fully isolated??
I also run a control cable to my transfer relay.

Frank,
On my 2el delta  loop for 75M, I tie all shields together at the switchbox.  I float it from the tower, but wud make little difference either way.   The f-b is between 20-30db from reports.  Never worse than 20db.   (So I know my radidio talks, caw mawn!)

T

I would think it would be more important when using an open stub.

Steve, Good Point after looking at your Hammond box. Maybe that is why the resonant frequency shifted so much. I guess I will try a longer stub to get back to short. Simulation showes reflector makes better FB. What is your stub length. Maybe I only need to be a few feet shorter. VSWR is so nice where I am.

The resonant frequency shifted (dropped) when I just had a shorted stub and no switch box. IIRC the stubs are around 18 feet. Varying the length of the stub will change the resonant point. But I didn't worry about that. I wanted the stub to only provide the needed amount of inductive loading to make a reflector. The progression to make the system was as follows.

1. Put up one loop and cut it to resonance at the desired frequency.

2. Take down the first loop and put up the second loop and cut it to resonance at the desired frequency.

3. With both loops up and spaced properly, add the stub length required to get the needed amount of inductive loading.

4. Trim both loops the exact same amount to achieve resonance system at the desired frequency.

5. Then check the SWR and see if any matching is required. In my case none was required. If it were, I would have dealt with it independent of the existing stubs (those would not have been touched). An L network or a series matching section would have been most likely applied.

18 feet is a good length for me. I could still reach the switch box from a small step ladder.  I'm about 20.5 feet now.
I'll have to drop one loop and see where it is at. It takes all of 5 minutes to lower an antenna.
I need to replace the connectors anyway. They are a weird old style form the LaPointe junk box. Tom, do you remember the parts crib free for all. I was told to help myself and everything is going.

Don't worry about the length of my stubs. They are irrelevant for your installation. Use as accurate a model as you can make to determine the required amount of inductance at the reflector loop feedpoint for your set up. Then use that formula I sent you to calculate the length of coax need to make the stub.

Yup, I want to simulate the loop and see what inductance it needs to resonate as a reflector. I have been letting the simulation tune the length of the stub and would like to bounce it off your method. Hopefully I can apply the changes this weekend. I think my mistake was making the stub a tad too long flipping it over. In my mind I have also looked at the stub as an echo delay line against the antenna.

If I ever get a day without rain, I might be able to install the switch box and true-up the loops here.  CQWW is coming up this weekend...might be fun to just work 40M.  I seem to do well with the present set-up anyway.  Been getting outstanding reports from EU.  Steve, believe it or not, I think Radio Shlock has those metal boxes like you used.  I wonder about the relays for full power.....could even use a couple of latching vacuum relays I picked-up some time ago.  Thanks for all the good info and discussion guys....been fun playing with these loop antennas.  I wish I had room for a set on 75 or even 160, but don't have enough real estate here.

73,  Jack, W9GT

Those relays are good for full power. Just don't hot switch them.   ;D  Remember, if everything is tuned up correctly, the impedance is around 50 Ohms, so the contacts don't need to be any larger than those of your TR relay. Ever looked at the size of the contacts in a Dow-Key?

Nice thing about SSB is you can hot switch them if you shut up for a second or so.

Steve,
Do I assume the paper you sent means X1 is inductive reactance in ohms and Fc is in MHz?? so about 5 meters of coax in your case of 2.33 uh.
So thinking ahead all I need to know is the inductive reactance needed to resonate the reflector at the lower frequency. I can't wait till my kid teaches me all the math I forgot.

Exactly. Just think of the stub as adding an actual inductor at the feedpoint.

I'm making Tom crazy so guess I should get on with it Saturday.
HPSDR is packaged so will have an accutate S meter again.

Yes, I agree the contacts should be fine, but the insulation might not be all that great. Might have to use something a little more robust, if you are using it for AM.  I have some latching vacuum relays that I got for another project and never used.  They might be nice also, since they can be switched with a pulse and then remove power.  I'm not sure though if it is politically correct to mount Jennings vacuum relays in a Radio Shack box.   :o ;D

73,  Jack, W9GT

LOL. Yes, the Jennings will reject the rad shack box - kinda like a kidney transplant.

I've used mine on full power AM and no problems. JJ has verified this too.

Steve,
I reduced the length of my stubs to 19 feet and turned the director into a reflector. I even saw the effect in simulation interesting how a slight length change has such a large effect. I think I got a little water in the transfer relay since the resonant frequeccy shifts 100 kHz down pointing west. When I had a director the resonant frequency only shifted 25 KHz. I didn't add the RG11 stub on the feed line. It is about 19 inches shorter now. Both loops 137 feet and simulation says 135.8. I had the relay in a plastic bag and there was some water in it this morning.  I'll let it sit a couple days before I mess with anything. One corner is down a bit and want to move the support to a higher tree this weekend. The  final relay box will be at about 15 feet and I ran 50 feet of LMR400 into the shack. 50 feet of LMR400 is very close to an electrical 1/2 wave on 40.
It seems to play and worked a couple guys on 40
Do you think I should reduce the stub or loop length to shift the resonant frequency up a bit?

The progression to make the system was as follows.

1. Put up one loop and cut it to resonance at the desired frequency.

2. Take down the first loop and put up the second loop and cut it to resonance at the desired frequency.

3. With both loops up and spaced properly, add the stub length required to get the needed amount of inductive loading.

4. Trim both loops the exact same amount to achieve resonance system at the desired frequency.

5. Then check the SWR and see if any matching is required. In my case none was required. If it were, I would have dealt with it independent of the existing stubs (those would not have been touched). An L network or a series matching section would have been most likely applied.

Steve,
I'm sweeping each loop keeping each side equal for wire length and stub. I do have 1 corner a bit lower than the rest of the antenna. This could be effecting the resonance when I swap the pattern. I was just wondering where you ended up with total wire length. My simulation says I'm a foot or so too long. I notice things changed when I pulled the relay up the tower. I think I was pulling the bottom in closer than 20 feet when the cables dropped to the ground.

I made it a little long and then took about three feet off each. So, IIRC, I wound up with about 138 feet of total length. I am using insulated #13 stranded wire.

I bet there is some water in the transfer relay since we are so close in lengths. I'm using #12 solid and insulated. The only difference is I have a tower going up the middle.
When facing East the resonant frequency is 7.13 MHz. Time to build the box I bought at Deerfield and mount it on the tower.

Success...finally. My simulation showed a slightly longer stub and slightly shorter loop then the HUZ configuration. Steve said something that made me think last night. He said you just have to start trimming the length until something happens. So this morning I went for it. I cut a foot off at a time on the west loop and after 2 feet the resonant frequency started coming up. I dialed it right to 7.160 and then the East loop to 7.160. The longer stub 18.8 feet of RG11 (.66) reduces the Q of the antenna enough that it is under 2:1 across the whole 40 meter band. FB looks good. now back to the afterburner.

The loops at sunset last night.

Click on the image below for a larger view.

No wonder you're Whirl-wide on 40M!   The lake, slope and swamp in front of it makes it play like it's at twice the height.    Dr. Love advised me not to get any more comparisons with you...

Getting the wire to stand out like that is FB.

T

Very cool picture.