Feedpoint impeadance of a 160m loop?

[kc6mcw]

What is the feedpoint impeadance of a full size 160m loop at 25' above the ground? And what would that impeadance be if operated on 80m? Is my thinking correct by saying a 160m loop is like having two half wave 160m wires end to end and fed in the center? Then you would be voltage feeding it right? And the impeadance would be on the order of around 2500 ohms or so? Why am I reading that the impeadance is 100 ohms? I dont get it....

[KF1Z]

I know a low 75 meter loop is about 100 ohms
At least... so says the Mighty-Fine-Junkbox, handy-dandy-swr-anal.izer.

[Steve - WB3HUZ]

Using exactly 1.06 wavelength for the loop size, and perfectly rectangular shape, the feedpoint impedance on 1.9 MHz in NEC2 models at 80 + j186 Ohms. Adjusting for zero reactance or resonance (making the loop slightly shorter) yields about 70 Ohms. Both of these were for "average" ground (rel. dielectric constant of 13 and conductivity of 0.0050 mho/meter).

On 80 Meters, say 3.8 MHz, the feedpoint Z is about  237 + j444 Ohms.

The Z on 160 will be much lower than your idea because its a complete loop, unlike your analogy where the ends are free.

[kc6mcw]

OK so if you fed this 160 loop with coax and wanted to multiband it, the swr on the line would only be around 1.4:1 on 160m and around 4.7:1 using it on 80m? Thats not much loss on the line due to the standing wave. Hmmm, a local ham near me tried this and his signal dropped from 40db over 9 using the V down to 20db over 9 on the loop. The coax from the loop went all the way to his Nye Viking tuner. I assumed the loss of signal was due to a large swr on the line. But 4.7:1 doesnt seem that bad though... is it?

[Steve - WB3HUZ]

That impedance would produce a SWR of like 20:1 on 50 Ohm coax. So, yes, the huge losses you saw are expected. If you want to multi-band the loop, you'll need to feed it with ladder line or open-wire line and a balanced tuner.

[kc6mcw]

So on 80m the feedpoint Z would be 237 + j444 ohms. So my math was 237/50 = 4.74:1 swr on the line. Am I doing this wrong? I believe I am...

[KA2QFX]

To answer the first part of your question, your thinking was logical regarding the antenna as a pair of end fed half waves, however, separate half waves would terminate in voltage nodes since they are not connected to anything, and logically the feedpoint would also be a voltage node (high Z).
 But by bringing the ends together in a loop forces the feedpoint to prefer (by feeding with low impedance line) being a current node, along with the opposite point in the loop. The voltage nodes now shift to the two points halfway around the loop from the feedpoint. 

As for the loss in signal I would attribute this more to cancellation of fields than to high SWR.  Loops by their natures tend to radiate predominately 90 degrees to the plane in which they lie. So any horizontal loop would radiate largely straight up, and cancel much of the signal toward the horizon. 

As for losses: If, as noted, the antenna is 237 + j444 the resultant impedance is 503 ohms. SWR = 10:1. 100 feet of RG-8 has a loss of less than 0.3 dB @ 75 meters. A 10:1 SWR will add an additional loss of only 1dB. So the losses you saw were NOT due to SWR, provided you were putting the same power into the line for both antennas.  Since you say you were matching with a Nye-Viking tuner I can assume power was roughly equivalent.   

There are other things to consider as well regarding pattern. Not the least of which is feedline radiation. Since operating any multi band antenna system over a wide range of impedances does not lend itself to the use of baluns, you should be using either a simple choke “balun” (ferrite beads on feedline) or open wire line.   Personally I would opt for the choke balun only because I prefer working with coaxial feed wherever possible. In this case it’s workable.

If you wish to pursue the loop you might try using a corner feed point and varying the aspect ratio of the loop from square to rectangular.  I would not make it narrower than 1/8 wave x 3/8 wave, but you never know what your situation might yeild.
Regards,
Mark

[kc6mcw]

Ok, I was assuming that you would be voltage feeding this antenna but why would I assume that? Hmmm, good point. So being a loop, it then shifts to a current node. Makes sense now why the feedpoint impeadance would be low. Good thing I had my thinking cap on today!

[Tom WA3KLR]

The SWR (50 Ohm system reference) for 237 + j444 Ohms is 21.54:1.

You can't just take the resistive part of the complex impedance and divide 50 into it.  Nor is the calculation the resulting ac impedance divided by 50 Ohms.

The formula for calculating the complex impedance SWR is slightly complex and the results is usually a much worse SWR than one would think.  (I learned this from Walt W2DU's book 18 years ago.) For those with a First Edition copy of "Reflections" the procedure is on page 23-5.  This is the Appendices Chapter - Appendix 3 - Determining SWR from R + jX, Method 2.

Example:   50 + j50 Ohms = 2.62:1 SWR !

[kc6mcw]

I am looking in my books for that formula. I didnt think it was that simple, but wasnt sure.

[Tom WA3KLR]

I will post an Excel sheet that does the calculation:

Below is a do-all spreadsheet I did many years ago to help with use of the MFJ-202 noise bridge.  It was for my own use only.  So at the risk of some confusion (no instructions), I am posting it here anyway. 

Since .xls files are not allowed, I have changed the extension to .doc .  Right click on file name and Save Target to disk and then rename extension back to .xls .

Your calculation example is done in the left column fourth set up from the bottom.

GL.

[Tom WA3KLR]

Joe,

I simulated the antenna with EZNEC 3.0 and came up with similar numbers to Steve WB3HUZ.  I adjusted the length to resonance very close to 1900 kHz, 131 feet on a side.

On 160, at resonance just below 1900 kHz = 85 Ohms.

Then on 75 meters:
3.6 MHz            194 - j316 Ohms    14.4:1 SWR
3.758 MHz (resonance) 198 Ohms     3.96:1
3.8 MHz              203 + j81 Ohms    4.74:1
4.0 MHZ            254 + j466 Ohms   22.3:1 SWR.

[kc6mcw]

What height above ground are you modeling it at? I am following along with my eznec. I have read appendix 3 and was succesfull with a couple of examples. I am slowly getting it...ha!

[Tom WA3KLR]

Important factors:

25 feet high
4 wires
#14 AWG wire, 16 segments per wire
source 50 % in from one side
copper wire loss
ground type = real/high accuracy
ground 0.003 S and 13 dielectric constant.

[kc6mcw]

Ok how would you factor in ground losses being that it is so low to the ground? Oh, and Im only allowed up to 20 segments total as I have the demo version! Hope this doesnt hurt it.

[Tom WA3KLR]

So you can only do 5 segments per wire then.  My program will tell you if you have too short or too long a wire segment.  I haven't re-run with 5 segments per wire.  This can change results slightly.

Ground losses/effects are taken care of well as far as I can tell in EZNEC 3.0 and later.  Nothing for you to do except have a simulator known to be accurate in this situation.  We've discussed the short-comings and pit-fall checks of simulators and antennas close to ground in a previous thread about a year ago.  Some handle this situation well, some don't. 

Accurate results are realized in the feedpoint impedance, pattern and gain numbers.  Programs that don't handle the "close to ground" situation show an unrealistic rising gain as the antenna is brought closer to ground, an obvious tip-off.

[Tom WA3KLR]

Joe,

I just ran a simulation with 5 segments per wire at 3.758 MHz.  No segment error message.

I picked a point on a pattern peak at 55 degrees elevation.
My original 16 seg. simulation showed 1.94 dBi gain and
feedpoint Z of 197.7 + j0.5477 Ohms.

With 5 segments per wire, that same pattern point has 1.88 dBi gain and
feedpoint Z of 211.2 - j2.953 Ohms.

[KA2QFX]

Thank you Tom,
You're right about the SWR. I used the Smith Chart at a glance and failed to moved my reactance origin to the right side of the chart. Duhh.
But even at 20:1 the losses due to SWR are still below 3dB. That was the point I was trying to make. 
Regards,
Mark

[kc6mcw]

Ok heres the deal. A local guy in my area put this loop up. Full size 160m at 25' above the ground. He is feeding it with 75 ohm coax. I am assuming RG-11. He is matching with a Nye Viking tuner. His signal was 40 over with his inverted V that he uses all the time. His signal dropped down to 20 over on the loop. The tested frequency was 3.8 mhz. We didnt test on any other freq. He is hoping for multi band operation. We are about 35 miles from each other. My first guess was large coax loss due to high SWR on the line...but my math shows this as a 14.49:1 swr on the line at around that frequency using Steve's quoted feedpoint Z at 237 + j444 at 3.8 mhz. That still doesnt sound like much coax loss. So why the huge signal drop? Could it be a lower takeoff angle? Eznec shows the signal going straight up just like the V. Could it be huge ground losses due to it being so low to the ground?

[Tom WA3KLR]

The radiation pattern of the the 160 meter full-wave loop at 3.8 MHZ (now 2 wave loop) is similar to a vertical, but not quite as severe.  You should be seeing a suckout of 20 dB at 90 degrees elevation.

For 2 stations on 3.8 MHz, 35 miles apart, the hop angle is very narrow, about 80 degrees elevation.  This would account for about 10 dB according to the pattern I get.  Perhaps cross-polarization is the rest of it.

First pdf is elevation plot.
Second pdf is azimuth plot at 80 degrees elevation. 
Note that depending on the azimuth between the 2 stations, the energy can be all vertical polarization to all horizontal polarization or a mix.

[Tom WA3KLR]

The gain of the loop is lower than the inverted V.

The gain of the 2 wave loop at all azimuths at 80 degrees elevation is from -4 dBi to -9 dBi.

The gain of a 75 meter inverted V (60' apex, 32' ends) at 80 degrees elevation is + 5.4 dBi, fairly circular pattern.

On the average, 12 dB lower gain on the loop, for the 35 mile hop situation.

[Steve - WB3HUZ]

Throw in a few more dB loss (possibly), depending on the type of tuner used.

[kc6mcw]

Ok, my math shows around 14.49:1 swr on the 75 coax line if used on 80m. Plus the tuner loss, plus the lower takeoff angle....all this might be the reason for the signal loss I noticed when he switched from the V to the loop. But 14.49:1 swr is around 3.1 db loss on the coax if my math is correct. Still doesnt seem like too much loss...? Around 25% power loss?

[Steve - WB3HUZ]

Yes it's only 3 dB. Another way to look at it is that if you are running the legal AM limit of 375 watts carrier (assuming 100% peak modulation), you are burning up 187.5 watts of power in your coax.

[kc6mcw]

Hmmmm, thats alot of carrier loss!