HFU HF Underground
Technical Topics => Equipment => Topic started by: ultravista on September 07, 2021, 1447 UTC
-
I built a 60 foot LOG yesterday using a DX Engineering BFS-1 at the feed point. The wire is solid core thin telephone cross-connect wire.
The LOG is very quiet and barely picks up distant signals. Compared to my Pixel 1B mag loop, this thing is practically dead.
I have watched numerous videos and read a lot about the LOG and mine is not as effective as others. This LOG is deaf on160 and 80 meters.
Any suggestions on what to do?
-
Interesting, at night, the LOG is different. I am picking up Israel 4XZ @ 4133 Khz with the LOG and Pixel. Also found conversations @ 160 and 80 meters. Not sure why day/night makes a difference.
-
Funny you should ask :) I have spent the past month or so working on a LOG for my LF/HF RX. Mine ended up 300 feet long, 75' per side square. I made my own matching transformer (?) , a simple 5:2 wound with magnet wire using a binocular core torrid as my feedlines are 75 ohm. It is incredibly quiet, but....I need to run a lot a gain sometimes (MOST times), depending on conditions. All I have to compare it to is (was) a 120' end feed longwire with a 9:1 balun, until a recent wind took it down.
The Longwire seemed to pick up more signals, but there was a lot of noise as well. The LOG signal levels are much lower, but I can really crank the gain with the low background noise. As with most antennas I ever bought, built and/or used...Conditions are everything. During the day on MW and HF it is pretty good, but after sunset it reallys seems to come alive. LF is all but dead during the day, but at night it's a totally different story.
My radios are a SDR and a Kenwood R1000, so nothing incredible there, but I do use a homemade pre-selector and preamp and it seems to help, again...depending on conditions
Ray Ray
-
The use of small AWG telephone wire may not be the best choice in antenna wire. This wire has a typical DC resistance of approx 98 Ohms per km and a characteristic impedance of 100 Ohms from 1 to 20 MHz.
Attenuation is in the range of:
2.6 dB/100m at 1 MHz;
5.6 dB/100m at 4 MHz;
8.5 dB/100m at 8 MHz;
9.8 dB/100m at 10 MHz; and
13.1 dB/100m at 16 MHz.
Unfortunately, the result being that the signals captured by the antenna are substantially attenuated by the length of telephone wire even before they have the opportunity to hit the receiver input.
-
Ray Ray
Nice to meet another Ray on the board.
-
~SIGINT~, thanks for the feedback. I wasn't aware that the small gauge wire could be the culprit. Interesting.
This is 60 feet. Is it enough to materially affect the signal?
What surprised me was the day/night testing. During the AM testing, the bands were quiet on the LOG but active on the Pixel. I commonly listen between 7-9 AM ET and get a lot of APAC stations. It was as if the LOG wasn't connected, it was so quiet.
The evening however was a different story. The same stations received on 160/80/40m by the Pixel were also received by the LOG. Albeit not as loud, the same stations were present, including Israeli CW @ 4133 Khz.
Day or Night, the LOG at 60 feet is dead to NDBs below the Broadcast AM band. The Pixel does a so-so job < 500 Khz and I can tune a NDB @ 415 Khz. The log received nothing.
-
What receiver are you using?
-
You are using a beverage transformer designed for 400-500 ohms terminating resistance. You LoG is a fractional wavelength antenna below the 20m band, or perhaps more around the 30m band due to resonance changing as approaching the lossy ground. Either way, your transformer is taking an already low-impedance antenna and lowering the ohms even further as frequency decreases, thus increasing mismatch and losses.
Start with removing the transformer IMO. Wind a 1:1 or whatever if needing common mode current isolation from the feedline, though as for balancing the LoG itself, the antenna already should be rather balanced regarding differential currents.
I only have a few snap-on ferrites on the coaxial feedline at the the feedoint my 148' LoG. I do have a decent RF choke back at the receiver end, but it is not really needed for my particular deployment, either. YMMV.
-
This is 60 feet. Is it enough to materially affect the signal?
Absolutely as you are dealing with signal amplitudes in the micro volts (uV) range which need to travel down the conductor to the feed point. The smaller the conductor the higher the resistance therefore you will get attenuation along the conductor. You also have less surface area to capture those waves.
Did you remove the insulation on the antenna wire? That as well has an attenuation factor. Then there is also the feedline loss. Every little bit makes a difference because all those little bits add up. Even the quality of the conductor can make a difference. Cheap mix alloy conductors with very little copper content make for poor antennae.
I also looked at the DXE-BFS-1 instructions and they specify:
Antenna wire: You will need a minimum of 3/8 wavelength of wire of any gauge between #8 and #20.
Your current wire is most likely 24 or 26 gage. 3/8 wavelength at 160 metres is 200+ feet of wire. Even at 1/4 wavelength you are looking at 137 feet of wire.
-
I don’t think you want to remove the insulation from the wire on a LoG antenna.
Although this isn’t the case for you since you just installed your LoG, both LoG and BoG antennas suffer performance degradation if they get subsumed into the ground. These antennas need to be sitting *on* on the ground, not under it, even a little bit.
-
What receiver are you using?
Chris, I am using a Yeasu FT-950.
RobRich, I also have the Balun One Nine (1:9) and a Palomar Engineers Bullet 9LF (9:1) UNUN. Would either of these two be suitable? Your 148' LoG, how did you arrive at 148 feet, trial and error or planned?
~SIGINT~, the wire is jacketed. I will replace the wire. I had no idea that cross-connect wire wasn't suitable for antennas.
-
I don’t think you want to remove the insulation from the wire on a LoG antenna.
Although this isn’t the case for you since you just installed your LoG, both LoG and BoG antennas suffer performance degradation if they get subsumed into the ground. These antennas need to be sitting *on* on the ground, not under it, even a little bit.
Agreed, removing the insulation would not be helpful :)
Regarding the type of wire, for a small antenna like this, 60 ft, the resistive loss from using the telephone wire may be trivial. I used unshielded CAT 5 (or was it CAT 6?) for my 1000 ft LoG, with all the conductors wired in parallel, seems to work fine as well. Granted, it's a *large* antenna, so that helps.
Which brings up this... signal levels from a small LoG are going to be quite low. If your receiver has a pre-amp, by all means use it.
Before building the 1000 ft LoG, I played around with smaller ones. The first was 50 ft or so, and I found to be be quite deaf, to be honest. To the point of not being remotely useful. I then went with a 350 ft LoG, that worked much better. Which is what led me to expanding it to 1000 ft.
-
A few gov and private party studies on buried hf antennas showed a dipole was efficient enough to be useful and certainly easier to deploy than a loop. They typically buried the antenna only a few inches below the surface, and used RG8 sans shield (meaning insulation and center conductor only) as the elements, waterproofing as needed. As you may have guessed these were EMP-related studies; can't be blown over if it's not above ground to begin with. Speaking of buried antennas, tests on hf within mine shafts showed some interesting properties, think it was a quartz mine so very dielectric, worked very well on 80 and 40, The ARRL antenna handbooks have a lot of these tests if you're interested
As for log types, special forces have long been searching for an antenna that was low profile with few showing any real promise. Those that do show some promise are fantastically expensive and are literally huge in area covered, at least for a wire antenna.
Here's one example;
http://www.hflink.com/antenna/elpa/
The Shirley and Jamaica Antenna has been suggested as workable;
http://arrl-ohio.org/SEC/nvis/nvis.pdf
As for hobbyist logs, I suspect ones with a single conductor, meaning no multiconductor phone wire, will be best, with as much wire on the ground as can be made to fit for reasons of gain. I suspect multiconductor wire will have odd capacitance effects and other undesireables as with using cat cable as an antenna - the pairs are specifically twisted so as not to interfere with each other signalwise.
-
I have posted this link before. This gives the background on why buried Log or Bog antennas don’t perform well. Figure 39 is worth looking at:
https://rudys.typepad.com/files/qexjul-aug-2016-bog.pdf
This should be no surprise if you think about the difference in performance of buried vs elevated radials on a vertical antenna. Verticals only need a few radials if the radials are elevated above ground, but they do need to be the right length. Buried radials require many of them to achieve similar performance but the specific length of them tends to be less critical as ground losses become more dominant.
A few years ago I put up (down?) a 1000 ft BoG for the winter DX season. I used a roll of cheap outside rated CAT5 UTP cable I bought off of Amazon. This cable was 24 ga copper clad aluminum (CCA) wire. I twisted all the wires together at each end and fed it with a 9:1 (I think) balun. The resistance of each of the individual wires was about 75 ohms, so it was probably about 9 ohms for the entire length of the antenna. An external preamp was not used.
I was pleased with how well it worked for LW NDB DXing. Part way through the season performance seemed to drop off a little, but I attributed that to conditions. I was still hearing many new stations.
When I rolled it up in the Spring I found that I actually had 4 pieces of wire. Some time during the Winter a rabbit or squirrel had chewed through the wire at multiple spots. My LoG had become a large OCF-dipole-on-the-ground!
One of my original reasons for measuring the wire resistance had been for future troubleshooting, but during the course of the season I never rechecked it. If I had it would have become obvious that the wire had been severed.
I will probably install another LoG this Fall. I may use copper wire this time but I really don’t expect a noticeable difference over CCA. Receive antenna currents are extremely small and (I squared R) losses aren’t very large. People build Beverage antennas using aluminum or Copperweld fence wire all the time. W8JI had some comments on wire types for Beverage antennas, but he said the biggest reasons to choose one over the other were largely focused on how well the wire survived environmental conditions and how difficult it was to work with.
http://www.w8ji.com/beverages.htm
There is probably a sweet spot regarding the minimum length of a LoG. Too little wire won’t capture enough signal. KK5JY suggests 15% of the target wavelength.
http://www.kk5jy.net/LoG/
My tests this Winter may use a roll of Home Depot THNN copper house wire (about 500 ft) instead of the 1000 ft CAT5 UTP as I want to see if this length will be useful at LF.
-
I suspect multiconductor wire will have odd capacitance effects and other undesireables as with using cat cable as an antenna - the pairs are specifically twisted so as not to interfere with each other signalwise.
The twisted wires depend on current flowing in equal but opposite directions in the wire pairs to achieve cancellation of radiated energy. For external induced EMI, the twisting ensures external fields induce equivalent voltages in both wires so that the induced energy can be negated by differential signaling.
Connecting all of the conductors together at each end negates these effects. It just becomes a lot of wires in parallel.
-
I suspect multiconductor wire will have odd capacitance effects and other undesireables as with using cat cable as an antenna - the pairs are specifically twisted so as not to interfere with each other signalwise.
The twisted wires depend on current flowing in equal but opposite directions in the wire pairs to achieve cancellation of radiated energy. For external induced EMI, the twisting ensures external fields induce equivalent voltages in both wires so that the induced energy can be negated by differential signaling.
Connecting all of the conductors together at each end negates these effects. It just becomes a lot of wires in parallel.
Yes, exactly ad NJQA noted, connecting in parallel negates this. You just end up with less resistance and inductance vs a single conductor of the same gauge, so it's a win win.
-
Size was not too critical to my design, but I was shooting for a loop circumference somewhat under the 40m band. It ended up being about 148' after laid out and deployed.
Here is the basic design of my LoG:
https://www.hfunderground.com/board/index.php/topic,29940.msg114696.html#msg114696
About a transformer, try it without one. Just connect the loop straight to your coax and see what happens.
Another consideration is knowing a LoG is mostly a skywave antenna. If you model a LoG, its primary lobe is vertically upwards, thus it can be a good solution for receiving via NVIS and high-angle signals. That is part of the reason your LoG might fair poorly during the day but much better at night, especially at low HF and below. Typically I use my LoG more as a regional antenna these days.
Yet another consideration is your soil conductivity. The characteristics of an antenna on the ground can vary considerably depending upon the actual ground. I have sandy soil with poor conductivity, which explains why my LoG performs about the same as when it was deployed years ago despite it now being mostly covered in shifting sand, leaves, etc.
-
I used unshielded CAT 5 (or was it CAT 6?) for my 1000 ft LoG, with all the conductors wired in parallel, seems to work fine as well. Granted, it's a *large* antenna, so that helps.
Chris, I have a large amount of CAT 5 cable. Please explain 'with all the conductors wired in parallel', meaning, all pairs connected together?
-
Yet another consideration is your soil conductivity
I am in North Georgia, the ground is red clay, very hard and compacted.
-
I used unshielded CAT 5 (or was it CAT 6?) for my 1000 ft LoG, with all the conductors wired in parallel, seems to work fine as well. Granted, it's a *large* antenna, so that helps.
Chris, I have a large amount of CAT 5 cable. Please explain 'with all the conductors wired in parallel', meaning, all pairs connected together?
Correct. I just stripped the insulation from all eight conductors one each end, and twisted them together. So it appears, electrically (DC anyway), as one conductor. Connected to a transformer and then coax to the shack.
-
Ground conductivity extremes can influence a LoG by like 20dB+, but your southern mountain soil is likely average to maybe slightly below average, so you should be able to disregard that issue.
https://forums.qrz.com/index.php?threads/loop-on-ground-antenna.622669/page-15#post-5742886
http://www.w8ji.com/receiving.htm
Your current LoG is basically *way down* on gain. Like others have noted, go larger for both the conductor and the loop circumference, though even then you might still need a preamp as frequency increases. Just part of dealing with a typical LoG antenna.
Also I would not worry too much about ordering or building an impedance transformer for the loop feedpoint for now. Start with a basic RF choke if you have *actual* common-mode noise. Get the LoG far enough away from local RFI sources, and you might need nothing but the coax between the loop and receiver.
http://www.karinya.net/g3txq/chokes/
-
I swapped the DXE Beverage transformer with the Nooelec Balun One Nine, connecting both wires to the Nooelec using the same thing cross connect wire. The change in transformer didn't seem to do much but I am having good results with the 60 foot LOG now.
It is less noisy than the Pixel and PA0RDT mini whip (both amplified) and is working surprisingly well. Perhaps there was something else going on on day 1 testing, but it is working now and in some respects, on par with the Pixel and Mini Whip. Time of day or other conditions must have had an influence.
The 60 foot diameter does nothing for NDB hunting < BC AM, it is completely dead there. Last night for example, I could pickup WWVB @ 60 kHz on the Mini (strongest) and Pixel (weakest) but not on the LOG. 60 feet may be too small. I am by no means an antenna expert, I dabble with antennas and only understand the basics.
I would like a LOG that is capable of pulling in signals < BC AM and would like know a suitable size to get there.
-
Start with a basic RF choke if you have *actual* common-mode noise.
Dumb question ... how do I determine if noise is *actual* common-mode noise?
-
My 148' LoG can receive longwave BCB stations with no preamp when propagation cooperates, but those stations are using extremely high-wattage transmitters. I am not really into NDB monitoring, but I have noted some when spinning the VFO down there.
An "ideal" start for a fractional loop is probably around 15% of the the desired wavelength.
Let us look at the bottom of the AM BCB band for an idea:
530KHz = ~565m * 0.15 = 277.5m
So around a 900ft loop there. Proximity to ground will lower resonance by a few percent, but we are dealing with such large loop sizes now, it is a "whatever" IMO.
If we disregard noise, ~1000ft+ would be a good start for an "ideal" low-frequency BoG. However, low frequencies are typically swamped by atmospheric noise, so you simply need an antenna with enough capture area for the receiver to detect the desired signal. Above that is basically just added gain, and it is about SNR instead of absolute gain.
TLDR? For longwave and lower, deploy the largest LoG you can. ;)
----------------
About common-mode noise, let us start with a very basic test assuming we are talking about a typical residential deployment. Tune your desired frequency with an antenna in the house, even if it is just a few feet of wire. Now hookup your outside antenna. Regardless of the signal level, do you hear the *same* noise? If so, you might have common-mode noise. If common mode is suspected, to further evaluate if the noise is being picked up by the antenna or the feedline, you would need to evaluate if the same noise is present at your antenna location. For example, by using a portable receiver.
----------------
Your Nooelec balun still is acting as a impedance transformer, though it is so small I am not sure what effective impedance transformation it is doing at such low frequencies. Hmmm. Considering a 60' loop is going to resemble almost an electrical short at longwave and lower, an 9:1 might actually work better if you turned it around at the antenna feedpoint. YMMV.
-
... turned it around at the antenna feedpoint. Not sure what you mean.
-
Your 60' loop has extremely low impedance at longwave and lower. Complex impedance aside, simple impedance likely is a few ohms at best.
You have a low impedance antenna, around 50-75-ohn coax, and probably around 50-ohm receiver. Your antenna is a significant mismatch, even to the point of almost appearing to your receiver as being *short* at low frequencies, thus incurring even greater losses for an already negative again LoG.
A 9:1 lowers the impedance even further by a ratio of.... well, 9 to 1. ;) Turning it around to become a 1:9, you can use the transformer to up the impedance match between the antenna and feedline/receiver, thus improving the match and incurring less losses at low frequencies. Your Nooelec balun will not have the right connectors if turned around, but you can overcome that with a little soldering.
Alternatively, the mismatch is part of the reason I said to try connecting the loop straight to the coax. Having no impedance transformation should be better than having a 9:1 transformation, at least at very low frequencies for your small LoG.
-
Got it, thank you.
-
I did scale the LOG to about 150 feet and irregularly shaped. Just what I had of a wire bundle, made to fit without cutting.
With a direct connect to 75 ohm coax, I am pickup a local NDB, 415 khz. It is faint, but visible in the waterfall. When connected with a 1:9 balun, the signal disappears. Moving back to my smaller loop, it too was able to pickup 415 khz, but much weaker than the larger loop.
The larger loop certainly helped, but it is strung around the backyard in an odd way, not sure I would leave this as-is since it doesn't appear to have much of an advantage over the smaller 60 foot loop.
The direct connect did make a difference.
The PA0RDT picks up WWVB @ 60 khz. The Pixel, LOG, and other antennas I've tried do not. I am receiving WWVB in North Georgia.
-
I've been using a 8x10m LOG for a few years w/ my AFEDRI-SDR and it's been a decent performer. It's using 12GA single conductor insulated wire. I do have to pump up the gain all the way on the SDR, however. I think i'm using a 9:1 balun at the moment.