(Dual-turn, 28-foot length wire loop prototype with 1:1 voltage balun.)
Last Update: 16 July 2005
As much as I like my gapped coax loops, I am also quite satisfied with small constant-current vertical loops made with larger-diameter wire or tubing. (Or using just the non-gapped braids of coax as pictured here.) They have similar performance to the gapped-coax loops, but require that you invest in a balun (stand-alone or inside the tuner) to help ensure that the feedline does not become part of the antenna. If you need to null local noise yet still be able to listen to most skywave signals, these loops really perform. To obtain the directional deep nulls, the overall circumference of the constant-current loop is typically 0.10 wavelength or less.
(2005 update:I've had great results when using parallel runs of coax feedline to the loop, [braids connected to each other at the bottom of the loop, braids also connected together at end of feedline but NOT grounded] and using the balanced voltage balun input inside the tuner. Then tape or zip-tie the coax together. In this rx-only application, balance is more important than strict impedance matching. You could also use twinlead or open-wire, but I have a very hard time keeping them balanced from other objects and upsetting the whole project.
(Click the image to enlarge)
The voltage-balun was essential to help me fight common-mode noise. See my balun notes below.
I initially chose 14 feet since my noise problem extends up into the 40 meter band; I didn't want the antenna to be longer than 1/10th wavelength because you start to lose your azimuth nulls with larger wavelengths of wire. I just did a quick calculation:
Generally, anything between .05 and .10 wavelength will work with acceptable efficiency and provide good nulls, (for small loops that is) and .08 wavelength is ideal. The value of 1005 is the classic textbook value for quad-loops, but you may want to research this value and see if 1032 or higher is more suitable. The next time I build my coax loops I think I'll try 1032 for 160-40 meters and 1050 for 20 meters and higher. (This is easy for me to remember if I just transpose the last two digits of the "textbook" value.) Maybe this is why I could still obtain pretty decent nulls when my loops started to exceed .10 wavelength!
Small loops perform well at their designed frequency as .10 wavelength circumference antennas, and also as .05 wavelength at half the designed frequency. Anything smaller and efficiency rapidly drops off. Anything larger than .15 seems to act very poorly and random as well until you reach a 1/2 wavelength or more.
Note that I have since opted to use 28 feet overall, because I wanted better sensitivity on 160 and 80 meters, and now at 40 meters the 28 feet of wire still gives me a slight null - adequate enough for me to null my local noise on 40. Unfortunately I don't have the room for a full-sized loop, so I had to wind it with half the normal diameter using two turns. See the EZNEC® antenna modeling plots below.
Here are some quick construction tips to get you up and running quickly. I'm still studying the antenna and will improve the page as time goes on.
I am running the loop right at the operating position. Here is what I'm using:
14 feet of #12 gauge wire formed into a loop. Coax braid (no gaps) is an option.
W2AU 1:1 voltage balun by Unadilla.
10 foot coax jumper from loop balun to unbalanced tuner input.
Common tee-type C-L-C antenna tuner.
The loop seems nearly omnidirectional for medium to high-angle skywave signals, (kind of an overhead oval) yet has great noise-nulling directivity at very low angles from 160 - 40 meters. These two qualities make it a great general purpose antenna especially indoors.
Let's take a look at the elevation angle for 40 meters. It shows good medium to high-angle skywave directionality. The other bands have much the same elevation pattern:
Small loops are good for short to medium-haul distances, while at the same time being capable of nulling noise via rotation. It is not a great low-angle DX antenna, although if you live with noise, you just might be able to hear them - albeit weakly and cleanly. THIS is where a preamp can help, but for general operations, you might not need one. DX'ers will probably want to run with both a tuner and a preamp.
Look at the azimuth angle for 80 meters. 160 and 40 meters are similar. Notice the deep null; great for nulling noise by rotating the loop:
This 14-foot circumferential loop is performing as a small directional loop on frequencies of 40 meters and lower in frequency, and as a somewhat poor omnidirectional intermediate-sized loop on bands higher than 40. In this case, I'd recommend building a smaller loop for the higher bands, since once you get larger than .10 wavelength in circumference, the loops lose their azimuth nulls.
The most efficient small HF loops are single-turn affairs. Multi-turn loops of this type are less efficient, but you may have no choice to wind a smaller loop with multi-turns if you can't find the space for a single-turn, such as with indoor applications. Personally, I never wind more than 4 turns.
If you are really space constricted, you could cut the dimensions down and run a 7-foot circumference loop. Just don't expect great performance on 160 or 80 meters. Rectangular loops might also be considered if you have a lot of vertical or horizontal space, but not much of both at the same point. Perhaps you have very high vaulted ceilings in which you can make long vertical runs for the sides of the loop whereas the horizontal runs would be much smaller.
Just remember that the key to small loop success is to enclose as much AREA as possible; keeping in mind that when your antenna starts to become longer than 0.10 wavelength in circumference, you'll start to lose the deep azimuth nulls.
The enclosed area is important since what we are trying to generate is differential voltages at the feedpoint, and a larger area between the wire of the loop results in larger voltages. Crunch your loop together to witness the loss of signal.
To help reduce the proximity effect of the turns, try to keep the multi-turn loop wires spaced at least one wire-diameter apart.
The use of large conductors, such as 1 inch copper tubing has the most benefit for 160 and 80 meters. If you have to make your low-band loop small, try to use the largest conductors you can find. For many of us, 1/4 to 1/2 inch tubing or ungapped coax braid does fine. At 40 meters or higher, it's not that critical - but in the end use whatever you have!
My real-world testing between single-turn and smaller multi-turn loops (2 - 4 turns) showed that multi-turn loops work very well. Even though I have initially reduced the sensitivity of the multi-turn loop by making the enclosed area smaller in diameter, the additional turn help to bring the sensitivity back in line. I don't think I'll ever make an 80 or 160 meter single-turn loop again - it was just too impractical indoors. On the other hand, I don't think I'll ever try to wind anything more than 4 turns on these bands. Single turn loops are still the best though...
These loops perform well very close to ground. In fact, if you mount the antenna much higher than .20 wavelength, no appreciable difference is made to low-angle reception, and for the most part you start to attenuate your high-angle directivity. As you approach 1/2 wavelength high, you start to get a REAL cloud receptor. So for me, I'll just keep my loops low. Mounting these small loops high away from noise kind of negates the purpose of local noise nulls anyway...
If you do decide to do this, be sure that you still have at least two nulls 180 degrees apart. If not, it is likely that your feedline has become part of the antenna turning it into a large vertical lollipop. Check for loop balance, use quality connectors, etc. My quick test is to place a clamp-on ferrite choke on the feedline - if any change in signal strength, directivity, or tuning is noted, there's a common-mode current problem causing an unbalanced situation. If the choke balances the system - great - but I prefer to find and fix the balance first.
On the left is a 40 meter .10 wavelength circumference loop mounted 1/4 wave high.
On the right is the same loop mounted up at 1/2 wavelength above ground.
I'm not sure these elevated heights are worth the effort - especially not at 1/2 wavelength elevation!
I have had the best results using a VOLTAGE balun at loop feedpoint or inside a tuner. Without it, the feedline became part of the antenna, and thus all I picked up was my local noise no matter what I did with the loop.
(I'm not condemning the use of choke or current-type baluns, it's just that the voltage-type balun seems to work better for me in this application.) Perhaps I need to find a higher-quality current balun, but so far the voltage baluns are doing ok.
Since I wanted my loop to be general-purpose and work across several bands, I made no attempt to match the loop impedance to the feedline. I'm just going to tune the whole system of loop and feedline. This may affect balun performance somewhat, but so far it is performing adequately with the loads presented by the loop in this rx-only application.
I really didn't want to use a balun, but found that I had to. I experimented with a direct connection to the coax without a balun, and got some directivity on 160 and 80 meters, but the common-mode cable ingress noise was pretty bad. It also changed my tuner settings radically since the feedline now became part of the antenna. Putting some clamp-on RF cable chokes on the feedline reduced the noise a bit but I still basically had a random-wire lollipop.
I reinstalled the voltage balun, took off the cable-chokes since they were no longer necessary, and got my deep nulls and quiet reception back! (Although I like to place chokes on all my feedlines as a general precaution. I hate common-mode current upsetting my antenna projects.)
Keep your balun connections neat and symetrical. For example, the voltage balun I use is a W2AU type and it has small jumper wires behind the strain-relief eyelets. Connect the loop wires to the jumpers close to the eyelets, and then symetrically dress the remainder of the balun leads neatly. I made the mistake of letting the balun wires hang in a hay-wire fashion, and attached the loop leads to random points along the balun jumpers. Although the antenna worked well, tighter nulls and better overall balance was achieved just by being a bit neater with my connections. It's worth the effort.
What about a 4:1 voltage balun? It works well! On a lark I thought I'd try a 4:1 voltage balun and see how much worse it would be. To my surprise, I still have my low-angle bidirectional directivity, my nulls are sharp all the way from 160 to 20 meters, and my tuner settings require about half the inductance! (except on 160 where my inductor settings stayed the same). It seems that as long as I can tune out the reactance of the antenna system, the loop-to-feedline mismatch isn't as much of a concern as I once thought on receive.
I had a 50-foot piece of coax left over from my earlier coax loop experiments, so I thought I'd experiment with it by using just the braid as the antenna element (continuous braid loop - no gaps). I had to take into consideration noise-nulling vs sensitivity for my location. I prefer single-turn loops, but in some cases I had to wind them into multi-turns (with one wire-diameter spacing) to fit indoors. In all cases I used my tuner to resonate the whole system. The lengths listed below are not super-critical.
(Update: I have come to prefer using real thick wire instead of coax braid as an antenna element, even if the gauge is thinner.) All these antennas can be used on other bands, just make a note that when they exceed .10 wavelength circumference, you lose the nulls -- that is until you reach half or full wavelength loops - which aren't really part of this project and are more "normal" antennas...
160 meters null optimized: 54 feet of wire
Had to mount this one outside as a quad loop. Great performance considering that it is close to .10 wavelength. I still need to find a good way to rotate it. I might wind this into multiple turns in a later project.
80 meters null optimized: 32 feet of wire
This is the best 160 meter loop I have used to date. Deep nulls on 160, medium nulls on 80, omnidirectional for 40m. Makes for a nice 1/2 wave loop on 20 meters and approx 1 wavelength loop for 10m (although for 20 meters and higher the loop can no longer be considered "small". This is a great all-purpose antenna.
40 meters null optimized: 16 feet of wire
Deep nulls on 80, medium nulls on 40. (I can still copy the locals on 160 ok, but since the loop is much smaller than 0.05 wavelength on 160, it's very inefficient.
20 meters null optimized: 8 feet of wire
This wire length now allows me to use a single loop of wire. However, I don't have a big noise problem on 20m that needs a null, so I favor a larger 1/2 wave gapped coax quad loop on this band and experimenting with gaps on the sides, or top corners. (But this is no longer a small loop project...)
I recommend using 1/4 to 1/2-inch diameter or larger conductor diameters for the loop if you desire them to be self-supporting. You can also use just the braid of RG-58 or RG-8 coax and affix it to a mast.
Small-gauges of wire don't perform as well as tubing does with loops under 0.10 circumferential wavelength, (especially on 160 and 80 meters) and larger conductor sizes, while offering greater sensitivity and a smaller bandwidth, may present a problem when considering the cost, weight, and general hassle of construction. The rule of thumb would be to use the largest diameter conductor that you find practical.
With these small loop antennas, going from #14 gauge wire to 1/2-inch copper tubing results in an approximate 10dB increase on 160 or 80. Changing from #14 to 1/2-inch copper tubing on 20 meters only results in approximately 2dB improvement.
(If you really don't want to turn knobs, you could just substitute a preamp .... )
Unlike a transmitting antenna, I'm making no effort to tune the loop to the feedline. I use a tuner to try and tune out the reactance of the whole system. Generally, If I can't get a loop to perform adequately with my tuner, I end up lengthening or shortening the feedline to make the tuner happy. Life's a bit easier in an rx-only situation...
Since small loops are high-q antennas, it is very easy to mis-tune or mistake a peak in your tuner settings for an optimal match of the system. If you have a noise-bridge, or antenna analyzer handy, this can make finding the right tuner settings much less of a chore. I'm going to describe doing it "by ear". Be sure to tune when the band is open or you are likely to miss the peak settings. Don't rely solely on weak local signals or weak noise to peak the loop during dead band periods. If you do, it's likely you'll trash this antenna before it has a chance to perform.
After building a new loop I usually get impatient and madly start adjusting the caps and inductor settings hoping that I can hear the peak quickly. Sometimes I get lucky, but more often than not, I dont' find any peaks, or I end up on a very inefficient one. Sadly, I resign myself to the fact that I'm going to have to do it in a more methodical fashion and maybe eat up an entire afternoon to find the settings for most of the bands.
Let's assume you have a typical C-L-C type tuner; a cap for the receiver side, an inductor, and a cap for the antenna side. You'll want to do this when the band is open, or perhaps tune to a local noise source. Here is my generic method going from the lowest to highest freqs (tedious to be sure, but I don't want to accidentally skip over a great match):
On the lower freqs, you may even want to advance the antenna cap settings by only half-steps until you find the peak. Ugh.
Even if you do find a nice peak, don't give up just yet! Make a note of the settings, and try it again with the next inductance value. You might be surprised at how well the NEW match point works. This is pretty tedious stuff, but the result are worth it - don't be tempted to be satisfied with the first peak you find!
After you've found the major match settings, you'll probably need to make slight capacitor repeak adjustments when you tune the receiver from one band edge to the other, especially on the lower 160 and 80 meter bands.
If you don't have a voltage balun handy and only have a bunch of spare coax lying around, check out my gapped coax-loop page.
since 9 March 2004