making a receive loop antenna

[nanZor]

Small vertical loop math revisited - 980

Like many other references, I've been harping on the 1/10th wavelength maximum for cutting your circumference of wire if you want the characteristics of a small vertical loop.

*** REMINDER - these are RX-ONLY loops directly connected ***

But what is the reference point you cut from mathematically? I have never actually seen it printed as a formula - just references to the .10 wavelength and then a project which seems to use less than the maximum desired.

Do you just double the classical half-wave dipole formula and use 936 / f mHz * 0.10 ??

Or do you use the one for cubical quads which uses 1005 ? Or my earlier reference to using 1000 ?

For years, I've been using these and getting good results, but the loops always seemed just a bit big after prolonged evaluation. As much as I hate to reduce my aperture area, I am now using loops just a tad smaller.

The best I could do was interpret the results from the ARRL Antenna Handbook where it describes a 20-foot circumference loop as being .037 wavelengths long at 1.81 mhz.

So, from that I am very happy with some new loops cut with:

980 / f mHz * 0.10

In fact, the ARRL tips off that while 0.10 is good, 0.085 is even better, since even though 0.10 wavelength loop is a current-node only type, the phase starts change above 0.085. I'll leave it at that, but now I think I have found something I can hang my loop-hat on:

980 / f Mhz * 0.085 = circumference in feet.

This is the formula I'm using and currently having good results with. Discernable difference? Can't tell yet since the great S/N ratio and no real lab equipment to measure rx-current makes it hard for me to say one way or the other.

 

[zz0468]

...But what is the reference point you cut from mathematically? I have never actually seen it printed as a formula - just references to the .10 wavelength and then a project which seems to use less than the maximum desired...

From everything I can gather, loop antenna design seems to be an imprecise science, so far as amateur and SWL antennas are concerned. All the authoritative references I can readily locate concerning loops provide formulas for describing the effective aperture, and the patterns of the E and H fields. So, it would be "easy" to work backwards and design a loop antenna with the desired characteristics. "Easy" being heavy on the math, and probably beyond the capability and interest of most amateurs.

Do you just double the classical half-wave dipole formula and use 936 / f mHz * 0.10 ??

Or do you use the one for cubical quads which uses 1005 ? Or my earlier reference to using 1000 ?

From all the construction references I can locate, the length of the conductor used to create the loop is 1/10 of one full wavelength. So, calculate the full wavelength, then divide by 10.

That gives the total length of the conductor, NOT necessarily the circumference of the loop, although in the case of a single turn, they would be one and the same.

Comparison with practical examples taken from things like ARRL Antenna Books of various vintages indicates somewhere someone used a velocity factor of about .71 somewhere along the line, but I don't know where that comes from. One classic example is for a 3.5 MHz loop, using 20 feet of wire. 3.5 MHz free space wavelength is about 280 feet, one tenth being 28 feet. Calculating with a velocity of 71% gives 20 feet, and that seems to be consistent with various examples of construction articles I found. That would be wound as multiple turns on a form 12-20 inches on a side, or 48-80 inches circumference.

For years, I've been using these and getting good results, but the loops always seemed just a bit big after prolonged evaluation. As much as I hate to reduce my aperture area, I am now using loops just a tad smaller.

The best I could do was interpret the results from the ARRL Antenna Handbook where it describes a 20-foot circumference loop as being .037 wavelengths long at 1.81 mhz.

The goal in a small loop is for current to have the same phase and amplitude throughout the loop. That's where the .1 (or .085, depending on source) wavelength or less comes in. The reason for keeping the physical circumference small is to avoid "antenna effect", where the loop acts more like a simple mass of metal, and exhibits none of the desirable loop properties, namely the pattern.

A single turn 10 meter loop would have less 'antenna effect' than a single turn 160 meter loop, so on lower frequencies, multiple turns might work better.

So, from that I am very happy with some new loops cut with:

980 / f mHz * 0.10

In fact, the ARRL tips off that while 0.10 is good, 0.085 is even better, since even though 0.10 wavelength loop is a current-node only type, the phase starts change above 0.085. I'll leave it at that, but now I think I have found something I can hang my loop-hat on:

980 / f Mhz * 0.085 = circumference in feet.

This is the formula I'm using and currently having good results with. Discernable difference? Can't tell yet since the great S/N ratio and no real lab equipment to measure rx-current makes it hard for me to say one way or the other.

If you read further in your book, you'll see where a small loop produces a very small output, compared to a dipole. The goal is to produce symmetrical lobes with deep nulls. So, your .1 wavelength of wire wrapped in a smaller circumference loop with multiple windings will minimize antenna effect that would distort the pattern.

 

[nanZor]

From everything I can gather, loop antenna design seems to be an imprecise science, so far as amateur and SWL antennas are concerned. All the authoritative references I can readily locate concerning loops provide formulas for describing the effective aperture, and the patterns of the E and H fields. So, it would be "easy" to work backwards and design a loop antenna with the desired characteristics. "Easy" being heavy on the math, and probably beyond the capability and interest of most amateurs.

I agree. That's why in this thread I'm concentrating mostly on the 1/10th circumferential length (in any shape), and getting the common-mode of the direct-connect coax under control to reduce antenna affect of the transmission line. I've more or less adopted the ARRL definition of 0.85, but the big issue for anyone interested is to get out of the books, put up a simple loop, and then go back to the books.

From all the construction references I can locate, the length of the conductor used to create the loop is 1/10 of one full wavelength. So, calculate the full wavelength, then divide by 10.

Yes, but do you caculate from 936, 1000, 1005, or some other value? (ie, 1000 / f Mhz * .10) A couple of miles of wire later in my own testing, about 980 seems to be the ticket - but it is not a showstopper if the loop is calculated differently. Build one, check it out, and trim or lengthen - the nulls will still be apparent that you've gotten yourself in the ballpark.

That gives the total length of the conductor, NOT necessarily the circumference of the loop, although in the case of a single turn, they would be one and the same.

True - although I'm not overly concerned with the shape - they can even be irregular vertically and still perform, although not asymetrically.

Comparison with practical examples taken from things like ARRL Antenna Books of various vintages indicates somewhere someone used a velocity factor of about .71 somewhere along the line, but I don't know where that comes from.

Possibly from the so-called shielded loops using coax for the antenna loop itself - which is wrong - when using coax, the common mode of the shield is the actual antenna (an rf clamp-on choke will prove it), and some mistakenly apply a velocity factor - but the VF ONLY applies to the differential mode on the inside, which is not where the actual antenna is due to skin effect. There probably is a VF from the jacket to consider, perhaps .95 or so, but it isn't critical for RX.

The goal in a small loop is for current to have the same phase and amplitude throughout the loop. That's where the .1 (or .085, depending on source) wavelength or less comes in. The reason for keeping the physical circumference small is to avoid "antenna effect", where the loop acts more like a simple mass of metal, and exhibits none of the desirable loop properties, namely the pattern.

You bring up a good point that I forgot to mention. For precisely this reason, as long as you have the transmission line choked off to prevent antenna affect, placement of the feedpoint doesn't necessarily have to be at the bottom - it can be anywhere really.

BUT, once the loop starts to exceed .10 wavelength in circumference, it becomes an "intermediate" sized loop - still current-node heavy, but it is starting to develop a voltage node which leads me to:

When using a .10 or smaller loop, if you decide to press it into service as a general purpose loop ABOVE .10, you may find that feeding it from one of the sides, (not the top!) will help bring down the elevation angle a bit. In fact just because of this thread I have been pleased with a 25-foot loop fed from the sides for reception ABOVE the .10 factor for general purpose listening where I don't have any noise and the loop is no longer considered a "small" vertical loop anymore. So far, so good for turning it into a dual-purpose antenna!

A single turn 10 meter loop would have less 'antenna effect' than a single turn 160 meter loop, so on lower frequencies, multiple turns might work better.

For rx-only, I've found that a single turn loop is just much less hassle than dealing with multiple turns spaced properly to avoid eddy currents, and the additional loss created by doubling the length of the wire. In this case, the aperture hasn't changed, so the directional properties are the same as a single-loop. But what I prefer may not be someone else's cup of tea. Try out multiple turns on HF - not my bag, but that shouldn't stop any reader from trying it.

If you read further in your book, you'll see where a small loop produces a very small output, compared to a dipole. The goal is to produce symmetrical lobes with deep nulls. So, your .1 wavelength of wire wrapped in a smaller circumference loop with multiple windings will minimize antenna effect that would distort the pattern.

Most definitely, but it broke my KISS principle. If one chases the S/N ratio, and covers up the S-meter, they might not find the need for a preamp immediately. But I will admit that a preamp can be very helpful at times. The standard preamp on most rigs seems to suffice for below 10mhz, and some rigs like the Icom R75 and others with twin switchable preamps are very small vertical-loop friendly!

When you model a small vertical loop, the great broadside nulls are very apparent, but so is the somewhat high elevation angle. I would say that it is similar to a dipole mounted only 1/4 wave high.

Someone asked me about my earlier statement about not mounting it high - "how high is high?". I'd say that mounting a small vertical loop higher than a 1/4 wavelength is a waste of time as the directional properties don't change that much. When you start to mount it even higher, you can get to a point where it becomes a cloud-burner. So putting it on top of a 50-foot tower (unless you need direct-wave reception, or physical separation from a noise source) isn't really necessary.
Mounting the typical small loop for 10mhz just above the house would be fine.

For anyone wanting to try a loop, stop reading and just cut a simple one and go from there. Small loops are almost a hobby unto themselves and there is a lot of fun to learn. We don't always get it perfect right off the bat. Maybe that's a good thing to keep the candle of experimentation alive.

 

[corbintechboy]

I hate to dig up an old thread but I have a question (rather then start a new one).

I just calculated the size of my loop based on the above calculation. I came to the number 20.825 in feet to have the loop centered on a frequency of 4Mhz.

With a variable capacitor, how high up the band can I expect an antenna of this size to perform?

 

[zz0468]

I hate to dig up an old thread but I have a question (rather then start a new one).

I just calculated the size of my loop based on the above calculation. I came to the number 20.825 in feet to have the loop centered on a frequency of 4Mhz.

With a variable capacitor, how high up the band can I expect an antenna of this size to perform?

Hard to say precisely, because things like construction technique and how/where it's installed will have some impact. In general, as you go higher in frequency, a loop will gradually start acting less and less like a loop, and more and more like just a hunk of metal. But I wouldn't be too surprised if an 80 meter loop turned out to be somewhat usable on 40 meters.

 

[nanZor]

I just calculated the size of my loop based on the above calculation. I came to the number 20.825 in feet to have the loop centered on a frequency of 4Mhz.

With a variable capacitor, how high up the band can I expect an antenna of this size to perform?

That's about the right maximum length of wire needed to center on 4 mhz. +/- a few feet either direction ok too.

About going UP the band. Your calculation is right (a few feet + / - ok, not too critical) to provide the largest antenna circumference and still have deep nulls - if you go too much higher in frequency on your receiver, (say 5.5 mhz and higher) the antenna is electrically larger than 1/10th wavelength, and can no longer be considered a "small loop" electrically. You will lose your nulls and the pattern will change as you go higher in frequency beyond what the loop was designed for. Roll back down to 4 mhz or lower, and the loop will be electrically "small" again, and have deep nulls.

You could use a capacitor to peak the loop at higher frequencies, but the physical size of the loop remains the same - you will lose your nulls regardless of peaking. Note that if the feedline is properly choked, and not too long, I just use a tuner at the shack end for convenience to tune the whole thing up as a system, rather than use a cap at the antenna. Purists won't mind running out to the yard every 50khz or so to tweak the cap.

Going much higher in frequency than the 1/10th wavelength that the loop was designed for might not be a bad thing if you don't need the nulling capability of the loop at those higher frequencies, or can just put up with the noise you may hear since you can no longer use rotational nulls.

 

[nanZor]

But I wouldn't be too surprised if an 80 meter loop turned out to be somewhat usable on 40 meters.

The caveat here is that yes, a loop cut to be about 1/10th wavelength on 80m will function well on 40m, (assuming you have tweaked the tuning circuit for 40m) but you will have lost your nulls since at 40m, the loop is no longer electrically small and behaves more like a "normal" antenna that now has both current AND voltage nodes.

The flip side is that if you cut a 1/10th wavelength loop for 40m, it will perform somewhat ok at 80m, since at 80m the loop is very much less than 1/10th wave, and the nulls will be very sharp and deep. BUT of course with the loop now skirting the practical 2:1 frequency ratio, the signal strength will be weaker overall due to the much smaller size of the loop electrically at 80m.

So it all depends on your practical requirements:

1) If I cut a loop close to 1/10th of a wavelength for maximal signal gain with good nulls, if I go higher in frequency on my receiver, can I live with the loss of the nulls since the antenna can no longer be considered electrically small?

or

2) If I cut a loop close to 1/10th of a wavelength, and go lower in frequency on my receiver, say at a maximum of 2:1, can I put up with the weaker signal strength due to the loop now being about 1/20th wavelength, even though my nulls are even better and deeper?

 

[nanZor]

I just calculated the size of my loop based on the above calculation. I came to the number 20.825 in feet to have the loop centered on a frequency of 4Mhz.

I just re-thought about your question and apologize for missing the part about "centering" on 4mhz, so I have a better answer than above.

Using the same equation, cut the loop for 5 mhz. With the 2:1 ratio, you can now go from 2.5 mHz to 5 mhz. At 2.5 mhz listening, the loop will be at the lower limit of sensitivity, yet have great nulls. At 5 mhz listening, this loop will be the most sensitive, but will be at the limit for having good nulls.

I don't know how much higher than 4mhz you want to go, but just keep in mind the 2:1 ratio of practical usability, and you can cut the loop for the highest freq you want to go within this ratio - assuming you still want nulls.