TheRadioBoard

Hope I'm not boring you all to tears with my loop antenna research. It seems some of you are interested in loss minimization, so I'll throw out another interesting tidbit I found.

The "ideal" inductor for a loop antenna is apparently comprised of several strands of separate wire formed into a tube:

http://www.eham.net/articles/26572 (see reply by AC5XP on August 22, 2011)

Pretty interesting information, but it would be mind-numbingly tedious to run all those wires along the loop surface. Perhaps long lengths of ribbon cable could reduce the work required.

That post also mentions that a too-large-diameter conductor also isn't good because the magnetic fieldline components start to undesirably become perpendicular to the loop surface.

The author mentions being able to use finite element analysis software to observe these phenomena. Anyone got any recommendations?

I'd suggest reading W8JI's comments. He is one of the smartest hams I know of. Do some research on him with a search engine. To save others the bother I'll quote him from that page:

I've mentioned in the past that a lot of hams put up some Antenna X, make contacts with it, then swear it's the best antenna since Marconi. There are similar comments in that thread. Yes, you probably can work people with the antenna described. You can work people with 3 feet of wire if you can make current flow in it. One key statement the author makes is that he "raised the radiation resistance." I'm more inclined to think he raised the ohmic resistance (as mentioned by W8JI), which would also explain the wider-than-expected bandwidth.

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

Yes, as I understand he designed the commercial MFJ loop antenna and took measurements of different designs (e.g. strap vs tube), so I tend to read closely what he has to say about magnetic loops.

That whole thread on eham has quite a bit of miscommunication, hurt feelings, polarization, and all-around noise, but if you take the time to read through all of it there is some very good information in there about magnetic loop antennas - including some information that I've not seen elsewhere (like the suggestion about a Litz-like tubular conductor for maximum possible loop Q). It sounds too tedious to make such a loop conductor (we're talking hundreds of meters of wire here, laid strand by strand!), but as I said maybe ribbon cable can be pressed into service here. It's a fascinating proposal, anyway.

Maybe a Litz-wire-like, flat, wide strip of many individual conductors (instead of the proposed tubular/toroidal multi-wire construction) would also work? That would allow using ribbon cable as-is. Any comments from the high-Q perfectionists here?

Not an expert in anything but I am wondering if computer ribbon was laid flat like your foil idea, would you still end up with the "edge effect" happening, when the wires are finally shorted together at each end?
It may in effect, electrically, be equivalent to a single flat conductor of similar dimensions?

(A bit off topic I know, re: your current question, but relevant to previous ruminations.)

............

Well there was some assertion in the eham thread that in the tubular case, the separate wires would prevent undesirable eddy currents on the surface, unlike a smooth tubular conductor. Perhaps the same is true for the planar case. I suppose modeling is the quickest way to get an answer. Much more study required.

In terms of inductors used in crystal sets, Litz has been found equal to or inferior to solid wire above about 1000 kHz. Don't mistake unfounded and untested speculation for "information." I went over and had a look at that guy's QRZ.com page. I'm glad he's having fun, but the bandwidth images he posted, apparently in support of his claims, are specious wastes of space, as several times pointed out by Tom (W8JI). A 3-foot loop is a 3-foot loop, and Tom told everyone what it takes to make it most efficient. There's no free lunch!

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

Not to mention that it would be a total waste of time anyway. The writer's understanding of skin effect and eddy currents is seriously flawed. The term "eddy current" has been a very unfortunate choice of words when used in describing skin effect, because it conveys a notion of electric currents swirling around randomly. This is definitely not the case. The so called eddy currents run in an arrow straight line longitudinally down the conductor, adding to or subracting from the primary current. They most definitely do not rotate around the circumference of the conductor as is implied by the construction of that antenna. At frequencies above where Litz peters out, the best possible shape for an isolated conductor is thin wall large diameter round tubing.



Same situation. The eddy currents are running longitudinally down the sheet. Separating it into isolated parallel strips would accomplish nothing except make a lot of work for yourself. In theory there might be some marginal improvement if the positions of the strips were occasionally transposed as is done with the conductors in Litz wire. However, I wouldn't expect to see any measureable effect.

BTW, while the principle that causes skin effect can be expressed quite elegantly, mathematically, that does very little to help visualize what is going on inside the conductor. The easiest to understand explanation that I've seen is by H.B. Dwight in chapter 10 (page 67) of his book:
Transmission line formulas for electrical engineers and engineering students
It's an old book, and some of the terminology is dated, but it's still very readable.

To avoid confusion it should be noted that the author of the helical loop antenna design is a separate individual from the author who proposed the tubular/toroidal multi-strand conductor, so there is no implied connection between the two ideas.


OK, how about a synthesized huge-diameter conductor, say, 10cm diameter tubing for a 1m-square loop? The huge diameter, impractical to manufacture and handle, could be synthesized by taking a large sheet of flat thin poster board, covering it with strips of copper foil tape, and soldering the seams together for a joined copper surface. Roll the board into a tube so it has 10cm diameter and 1m width and solder the tube edges together. Construct 4 such tubes, cut each tube's ends at 45 degrees, and solder them together for a 1m-square loop.

First off, would the cardboard-like poster board backing on the interior of the tube cause losses?

Second, is 10cm diameter excessive? Again to quote from the eham thread:

http://www.eham.net/articles/26572

Thanks for the all the discussion guys. It's very informative.

Right. I'd started to comment on the tubular/toroidal multi-strand conductor antenna comment before actually checking the entire eham thread, and meant to preface my comments appropriately, but missed doing it.



I think this is a fair warning, although I would offer a different reason for it. Most of the scientific analysis of inductors, loop antennas etc. makes the assumption that the conductor diameter is small compared to the loop diameter. This makes the analysis much simpler, because you can use one or two dimensional methods rather than 3 dimensional methods which get very difficult very quickly. It doesn't necessarily mean that if you break the rule that you'll get poorer performance, it just means that it's more difficult to analyze. FYI, most finite element analysis approaches, that I'm familiar with, will simplify a 3D model to a 2D model in order to lighten the computing load.

If you're determined to construct an antenna with metal foil, then I think that forming it into a tube is a more practical idea than keeping it as a flat sheet.

Just wondering about the "corners". It might not be a problem at your frequency, but ... would HF "wave guide" technology come into play here? (I think there is a minimum radius WRT frequency, for corners)

............................

In that vein, here's one design that looks interesting, using aluminum foil wrapped around a thicker aluminum duct:

http://www.alexloop.com/artigo24.html
http://hps.infolink.com.br/py1ahd/gallery24.htm

I have no idea how to estimate the losses here... the circumference of the foil wrapping (where it wraps back on itself) can't be soldered along the seam, so we have an imperfect lossy seam; the overall surface isn't flat and smooth; there's unknown influence of the inner duct tubing; losses are incurred with the mechanical connection to the variable capacitor.

That said, the designer uses an FT-817, same as my rig, so maybe it's worth a shot. I might leave some loose foil flaps at the ends of the tube to serve as the capacitor.

Interestingly the designer indicates that the original corrugated aluminum tubing could not be resonated, but wrapping it with foil solved the problem. Any explanation for why the original tubing might not resonate?

There are formulas available that will allow you to calculate the radiation resistance of a loop antenna, and there are also skin effect formulas available to calculate the AC resistance of the loop. This AC resistance (plus other misc. resistive losses) will be your antenna losses. You want this to be much lower than the radiation resistance.

For example, I recently built a Part 15 AM transmitter with a 3 meter wire antenna. Using the formula for an electrically small monopole antenna, I get a radiation resistance of 0.09 ohms at 1500kHz, which is quite terrible, but there's nothing you can do about it, because the rules limit you to a 3 meter antenna. Calculating the resistive losses of the 3 meter wire (#14 AWG in my case), I get 0.06 ohms. So, even if there are no other losses in the system, nearly half the power is dissipated in resistive losses. Adding in the other resistive losses, it's likely closer to 99%. This is the problem with electrically small antennas: the radiation resistance is so low that it is usually swamped by all the various losses. If you're using it for receiving only, then you can compensate for these losses with amplification, but for transmitting, there's no remedy other than to minimize ohmic losses in the antenna/ground system.

Flexible aluminum ducting is cheap and readily available here. I can get a 3m length (i.e. 1m loop diameter) with a cross-sectional diameter of either 10cm or even a whopping 15cm.

Any thoughts - for a 1m diameter loop, is 15cm getting too thick?

If the efficiency calculators are to be believed, such a loop has a theoretical efficiency of 49% on 40m! Not bad. Obviously that will be impossible to achieve in practice with the aluminum not being solderable (necessitating a mechanical connection to the capacitor, whatever form it takes).

This is all assuming I will be able to resonate the ducting at all - as mentioned above, another designer had to wrap the ducting in foil for resonance, for reasons unclear to me.

EDIT: Another benefit of huge-diameter tubing is that the surface area of the loop surface itself may be enough to form the capacitor plates, with a single common conductor (e.g. flat plate) laid over both ends to complete the "split stator" capacitor assembly. That would eliminate loss through mechanical capacitor connections to the aluminum tubing. I also have an idea brewing for how to make this remotely tunable via a repurposed electric screwdriver.

Another vote of confidence for strip-based magnetic loops with overlapping ends to form the capacitor: the DL7SAL loop (note: linked text is in German). http://www.qrpforum.de/attachment.php?attachmentid=4662

What I've read so far indicates people have achieved good performance and high-Q with them.

Interestingly one guy accidentally ruined his Q by attaching a rubber insulation along the edge of his strip. After removing the rubber, the Q significantly improved:
http://www.qrpforum.de/index.php?page=T ... eadID=6591

So for my idea of using copper tape or aluminum foil, the choice of backing material and surrounding structure to support the foil needs careful thought to avoid killing the Q. Maybe polyethylene tape, if I can get a hold of any. But there's still the conductivity and dielectric loss of the adhesive to consider, which is probably impossible to find out.

Just thought I'd add that aluminum is solderable, with some caveats.
You'll have a problem if it's to thin.
To solder aluminum you need a soldering gun, Aluminum wicks the heat away
quickly. First make a puddle of solder, then in this puddle (which actually shields the aluminum from reacting with oxygen) using the soldering guns tip, scratch the surface (to remove the oxidized layer) continue scratching and you will find the solder does stick to the aluminum. Once you get a layer of solder (basically you have now tinned the aluminum) you can solder a wire to the tinned area. (the puddle).
Mikek