1/8 wave spacing and dual band

[WA1GFZ]

A couple questions.
I was going to shoot a couple lines in the woods today but there was a bit too much breeze.
Then I decided to cut down the two element array in the back yard for 40 meters. I first ran a rope between the top of the tower and a 80 foot tree. Then I attached the two elements 32.5 feet apart. I was about to cut the elements when I stopped and decided to give close spacing on 75 meters a try.
So my first question would be outside of driving the input Z lower anybody ever try this on 75 or have an opinion?
Then I thought, Why not dual elements and use it on 75 and 40 meters. 40 would be 1/4 wave spacing.
Any ideas? gfz

[Steve - WB3HUZ]

According to UN4UN, FB results are possible with a 10 meter/33 foot spacing. The Z is around 20 some Ohms but easily matched. If you could switch in a second matching network for 40 you would have a twofer. Cool!

[K1JJ]

Generally, 30 ohms is the lowest you want to go with a Yagi (or most arrays) before the losses start eating up the hard-earned gains. This is assuming aluminum tubing. For wire, be sure to use at least #10.  I wud expect a good 0.6db loss at 20 ohms or less. More than 2els = even greater losses. Considering the forward gain is only about 4db, this represents a big hit.

At 60' apex height, I assume you will be using inv vees and not loops or the avg height will kill any TO angle on the loops.  Even so, the avg height of inv vees on 75M will probably be down around 50'? (in your back yard) so this is less than 1/4 wave above ground. 40M will be better, but still less than 1/2 wave length.

The other extreme is using a design like the 3el 40M wire Yagi I posted today. It uses a 60' boom. The input impedance is about 50 ohms (low loss) and very broad coverage on 40M. The gain compromise in pattern is only about 1/2 db.  But I realize you may not want 30' between els on 40M for support reasons.

So, bottom line is, I see two challenges of potential IR losses and maintaining a TO angle that will give good DX capability.  Being aware of these items shud direct the design.  Make a BIG effort to pull those ends up as high as possible - and keep the input impedance up to a reasonable value for wire..

Still, its WELL worth the effort, whatever you end up with.

T

[WA1GFZ]

Yup,
Simulation shows about 18 ohms on 75. Pretty windy this morning so may work on the phasing today.  I saw some nice front to back results in Europe against the reference dipole. I just have hardwired JS lengths of coax coiled up in the basement. Time for a transfer relay or switch.
I wonder if I went to folded dipoles on 75???
Tom, the last time you were here there was a rope way up in a tree. That is the west support. There is a better branch but I'll never hit it with wind blowing.
Wire is #10 solid

[WU2D]

Frank,

I know this sounds silly but what about hanging two trap verticals from the rope. I had very good low angle DX results with my 80M vertical wire Yagi. If you drove the 2 elements you might get some interesting results. Again, I would start with two 60 ft long verticals at the 33 ft spacing and a few elevated radials for testing.

Tom can you model this?

Mike WU2D

[K1JJ]

Hi Mike,

That system would work FB and some use it on 40M these days with good results.

I think the "IF" factor is the ground conductivity out to a mile or so from Frank's house. Also, having been at his house, his house would be very close to the verticals and kill any chance of taking advantage of the low angle energy. The tower and close guys lines would also be a challenge to make sure they were not interacting. 

That said, as far as modeling, my old miniNec DOS does a poor job simulating grounds - just uses a mirror surface, so wud be useless for that modeling. It's good for horizontals or perfect groundplane verticals. Maybe the other guys could model it with the later software.

Frank - just to give you a head's up ...18 ohms is really very low and will show up as added forward gain loss. A f-b is easy to get.... even a high loss antenna like a beverage has a great f-b, but has negative forward gain.  Just be aware that the only way to really find out if it has forward gain is to use a reference dipole at the SAME height and orientation. You will have to look closely to see 4db gain and probably less with an 18 ohm input impedance.  Even using 3" aluminum tubing tapered down to 3/4" shows significant losses at 18 ohms. You won't find any commercial Yagis that are that low. Copper wire will be even higher than the tubing. 

Though, your acquired receiving f-b  advantage will still outweigh any single dipole you now use, so is worth it if you go that route.

T

[WA1GFZ]

Tom,
Interesting I did see the F/B against the EA3 last night but little change in signal strength. The reference dipole is presently higher and facing a bit more to the north. Also over the stream.  The noise from the cold front coming dropped a good 10 dB on the spectrum display
butt hurts sitting at the computer driving the simulator past few hours. The coax lengths I have I will be setting up phasing later. going to play in the garden for a while.
The S/N advantage looks very usable.
Mike vertical would be a pain in the middle of the yard. I could string some vertical elements with aluminum tubing along the property line but would need to bury radials. That is an option 33  foot radiator is no big deal and I have the tubing.   

[Steve - WB3HUZ]

20 Ohms on 80 meters may not necessarily induce too much loss. With a 33 foot spacing, one element has a feedpoint impedance of 17.5 + j42 Ohms. You only get 0.6 dB loss using 100 feet of 9913. That's only 0.23 dB more loss than using RG-8 on an antenna with a perfect 50 Ohm impedance. Use hardline and the loss only 0.12 dB - that's LESS than the "perfect" antenna fed with RG-8.

[K1JJ]

20 Ohms on 80 meters may not necessarily induce too much loss. With a 33 foot spacing, one element has a feedpoint impedance of 17.5 + j42 Ohms. You only get 0.6 dB loss using 100 feet of 9913. That's only 0.23 dB more loss than using RG-8 on an antenna with a perfect 50 Ohm impedance. Use hardline and the loss only 0.12 dB - that's LESS than the "perfect" antenna fed with RG-8.

I'm talking about the wire loss in the elements. You can select copper wire and a wire gauge in modeling and it will show you the db losses. . I'd have to model it on 40M, but on 75M, I would always hesitate to build a Yagi that was lower than 30 ohms cuz of the increased wire or tubing losses. The pattern looked much sharper and showed more gain, but until super conductors become available we're somewhat limited.  A  10 ohm Yagi shows tremendous gain... :-)


As far as coax loss... I wud definately match that 17 ohms to the 50 ohm coax at the element feedpoint. Then we're back to the lovely 1:1  0.12db loss per 100'.

Anyway, I look at it as a percentage loss vs: forward gain.  If  1 db is lost in a real world situation using 200' of coax and #12 copper wire, then 25% of the forward gain is gone into heat.  It's only a db, but then that leaves only 3db of forward gain for all that effort.

** I just modeled a 40M 2el Yagi with a 20 ohm input impedance. Using #12 wire, the wire loss was  0.46 db.  Add some more elements and the loss goes higher.

T

[Steve - WB3HUZ]

Excellent point. I gots to remember that wire loss. I guess that's why they used folded elements or tubing for the close-spaced W8JK arrays back in the olden days.

[K1JJ]

Excellent point. I gots to remember that wire loss. I guess that's why they used folded elements or tubing for the close-spaced W8JK arrays back in the olden days.

That's something that I need clarification on....

I can see using the folded els to bring the impedance up to match to a convenient feedline value.  But how about the element wire losses?

Lets say the input impedance of a 5 element Yagi is 20 ohms and the losses through ALL the wire elements is 0.6db.  If we make just the driven element a folded element (20 ohms stepped up to 80 ohms, 4:1) what happens to that wire loss? Won't the parasitic elements still contribute to this IR loss?

 I don't think the radiation resistance is affected at all by a higher impedance driven element.... but how does that relate to the driven el impedance step up vs: overall wire losses?  Just for driven els only?


I've not used folded elements nor pursued reading about it lately. Any ideas here?

Frank/GFZ was axing about using folded els for his 2el driven array as a step-up to 50 ohm cable.

T

[WA1GFZ]

Well,
I got the reference dipole a bit higher FP is about 60 and the ends close to 80.
I played with the phasing. The best F/B was a bit over 10 dB but gain not great with the ends close to the ground.
The Elements are still a bit long so maybe I should do some testing in the CW band.
Next I'm going to simulate folded dipole elements to see how the match looks. The coax runs are RG 213 and RG214 . 141 and 95 feet.

[WA1GFZ]

Tom I was looking at folded elements to get a better match to the 50 ohm coax

[K1JJ]

The Elements are still a bit long so maybe I should do some testing in the CW band.
Next I'm going to simulate folded dipole elements to see how the match looks. The coax runs are RG 213 and RG214 . 141 and 95 feet.

HI Frank,

I found that the inv vees need to be shortened by about 2.7% - at least in my case at 80' apex and 65' end heights.   Huz is experimenting with his array and found his too, was resonating down around 6.9 mhz... but he may have a different cause using loops and stubs.

Yes, forward gain is a bear to achieve. Many of the old multiband trap 10-20M Yagis had such large losses they had unity gain on 10M vs a dipole. The users loved to see the nice f-b when turned, but had no clue to what was really going on...  

Using your reference dipole, you will see the real story and not fool yourself.

Let us know your findings with the folded dipoles. Be sure to toggle on the wire losses and see if they change - that's what it's all about, trying to minimize loss any way possible to maximize forward gain.  (f-b is easy to get)

T

[KA1ZGC]

Well,
I got the reference dipole a bit higher FP is about 60 and the ends close to 80.
I played with the phasing. The best F/B was a bit over 10 dB but gain not great with the ends close to the ground.
The Elements are still a bit long so maybe I should do some testing in the CW band.
Next I'm going to simulate folded dipole elements to see how the match looks. The coax runs are RG 213 and RG214 . 141 and 95 feet.

I'd really like to run that some more, Frank. The difference I saw was mostly in levels of QSB. One ant was more stable than the other. Unfortunately, the more you switched back-and-forth, the less conclusive the results were, but there was a definate difference in signal stability, even if not in the guess-meter reading.

We were in late afternoon/early evening conditions when everything was in flux. It would be interesting to try it again in broad daylight. You've always put in a good afternoon signal here.

Next time I hear you on, I'll probably accost you for another test. You've been warned!

[KM1H]

Yes, forward gain is a bear to achieve. Many of the old multiband trap 10-20M Yagis had such large losses they had unity gain on 10M vs a dipole. The users loved to see the nice f-b when turned, but had no clue to what was really going on... 

How very true. I found that out the hard way in the 60's with a TH6. A friend about 2 miles away had a 3 el HB 10M monobander at the same height and was blowing me away every time in every direction. That was my last trapped antenna.

As far as the thread project, would dual driven elements in a log cell plus a parasitic reflector take care of the feed impedance? It seemed to work well with the old KLM yagis. My old 4el 40M KLM worked great when it stayed up.

Crashcraft also appears to have excellent efficiency with their 2el 40M.

Carl
KM1H

[WA1GFZ]

I'm not seeing much gain but F/B is pretty good. Presently the reference antenna is up  higher but I do see a difference in the BB noise floor. So if the signal level is the same between both antennas but the phase antenna has a lower noise floor this is a good thing. The SDR display is cool to monitor multiple signals at the same time as antennas are switched.
Yesterday I moved the reference yet again and there may be a way to sneak another element at the same height. The spacing will not be as close. I found 42 feet is a nice spacing.
Tom, I know the inverted Vee needs to be shorter but holding off with the trim until I settle on a configuration. Simulation says the elements want to be about 58 feet rather than 61.5.
Yes, Log cell is cool  but hard to reverse. I never have or ever will run with traps.

[K1JJ]

I'm not seeing much gain but F/B is pretty good. Presently the reference antenna is up  higher but I do see a difference in the BB noise floor. So if the signal level is the same between both antennas but the phase antenna has a lower noise floor this is a good thing.

Frank,

Yesterday I noticed that the NE 40M Yagi was at least 10-15db quieter than the west Yagi. This is ambient background noise, not static crashes during the afternoon.  I've noticed this for a long time now, even on 75M.   It appears that maybe the composite integrated noise sources to the west (higher population) than to the east (RI and then the ocean) are the culprits.

So keep this in mind when you make f-b noise measurements. ie, When you switch your array to the west, don't be fooled into thinking something is wrong with the f-b cuz the noise level came up.  Just listen to real on-air signals to get the best readings on f-b.

In the past, whenever I needed to tune my f-b, I'd tune in a group of ssb alligators chatting from the southeast. The would be the perfect beacons to null out when tuning els or simply making rough A/B measurements.

BTW, Not seeing any gain, huh?...  probably due to the very low impedance (losses) and the ends being almost on the ground. On 75M, an average height of at LEAST 60' is needed to make it start to come to life.  Your house could also be blocking the NE at that low height?   Get the ends way up and increase the input impedance and it may become a whole different antenna. 

T

[WU2D]

The folded element idea sounds good for 80M roughly quadrupling the impedance.
Operating the same dipole on the second harmonic - 40M, is another story because the full wave folded dipole ends up looking like two half wave shorted stubs so I would imagine that the Z goes way up? Tom?

I have to give Tom something to explain.

Also take a look at figure 2 and the twist on the folded dipole.

Mike WU2D  

[WA1GFZ]

Tom it is very cool to watch the spectrum display and see some signals come up while others go down. Yes the early PM noise to the West is interesting to see and I thought I had a problem when phased to the East.
Gain. The reference dipole is much higher on the ends, close to 80 feet and center over the stream bed. The dipole is much more efficient compared to the inverted vees. 

[WA1GFZ]

Well, 32 feet spacing is making me crazy so moving out to 40 feet. This looks a lot better in simulation and still good on 40m.
Front to back doesn't work all that well even if I drive it like a log cell. gfz

[WA1GFZ]

Well 40 feet spacing dialed into the Southwest. Most of the local NE signals down about 10 dB from the reference antenna. I just heard a guy in PA. up at least 6 dB over the reference.
40 feet spacing seems a lot easier to deal with so time to do the phasing and see what it does when the pattern is switched. Simulation likes about 140 degrees phasing.

[K1JJ]

Well 40 feet spacing dialed into the Southwest. Most of the local NE signals down about 10 dB from the reference antenna. I just heard a guy in PA. up at least 6 dB over the reference.
40 feet spacing seems a lot easier to deal with so time to do the phasing and see what it does when the pattern is switched. Simulation likes about 140 degrees phasing.

Yep, I've always tried for 40' between elements on any array for 75M.  The big gun Yagis use 80-90' total for the 3el rotarys. The more compact ones use 60' total for 3els.   So at 40' for 2el, you're in good shape.

You will have a better band coverage and less IR losses with the higher input impedance.

Get those ends up as high as possible and you'll be golden.


T

[WA1GFZ]

Ends are the next step. I might move them out further to get more height.
Cool that it fits in the back yard so I don't need to fight the woods.

[Steve - WB3HUZ]

I modeled two phased-arrays and included wire loss. There seems to be little difference regarding IR losses between an array spaced 33 feet or 43 feet when using #12 copper wire. The feedpoint impedances are higher on the wider spaced array (not surprising) and probably a little easier to deal with but the closer spaced array still seems reasonable.

Parameters for both arrays

Height: 70 feet
Ground: average (rel. dielectric constant of 13 and conductivity of 0.0050 mho/meter)

Spaced 33 feet

Z1: 21 + j38
Z2: 27 - j53

Take-off angle (TO): 42 degrees
Phase Angle: 145 degrees
Peak gain:  9.49 dBi
F/B at T/O angle is about 38 dB


Spaced 43 feet

Z1: 40 + j52
Z2: 33 - j51

Take-off angle (TO): 42 degrees
Phase Angle: 140 degrees
Peak gain:  9.67 dBi
F/B at T/O angle is about 40 dB