I am experimenting with 3D "fractal" antennae, made from copper tubing The object is to determine their resonant frequencies by manually sweeping my 20MHz function generator (sinewaves) and watching for "peaks" on a digital CRO.
Unfortunately, I am unable to detect the hoped for response.
The full wavelength of 20MHz, the maximum limit of by generator, is 15 metres. My antennae are sized down around the 5th and 6th lower harmonic of that. Being "fractal", I had hoped this would not be a problem. Maybe it is. Or I should be using a different wave form, higher voltage spike, etc.
I am trying to work within the limits of what I have. Is there any solution that does not involve buying a new function generator and CRO?
--
John Larkin Highland Technology, Inc
jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com
Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro acquisition and simulation
Peaks in what? If it were me, I think I'd be looking for dips in return loss.
My antennae are sized down around the 5th and 6th lower
Are you saying that you took someone's theory and designed an antenna that would have resonances below 20 MHz?
"Sized down by a factor of 6" sounds like a recipe for failure???
I'm confused by "sized down" and "lower harmonic" Are you saying the thing is 90-meters big? Or that it's 3-meters big?? Or what? Neither sounds right, but the 90-meter one should let you see some peaks and valleys with your function generator.
Being "fractal", I had hoped this would not be a
Using the "fractal" buzz word doesn't make "hope" a good design parameter. Electrons and fields don't care much about hope ;-) What did the design equations say? And did you build that?
Maybe it is. Or I should be using a different wave form,
It's theoretically possible to do it in the time domain. I think you'll spend less money overall if you stick with frequency until you get a lot more experience.
As I see it, you have two choices.
1)Build an antenna that fits the tools you have.
2)Buy tools that fit the antenna you have. Ain't no magic bullet.
I's start by convincing myself that scaling the thing by a factor of 6 is a good idea.
A grid dip meter is one place to go. The older units that actually have a grid are often way better at this than the newer ones. Old hams have old grid dip meters in the basement. Maybe you can borrow one.
A balanced push pull frequency doubler will generate even harmonics (mainly the second) and suppressing the fundamental quite well. This would extend your frequency range to 40 MHz. Some filtering and amplification and a second frequency doubler and you are up to 80 MHz.
Not really surprising. You would be better off making an oscillator that excites the array at or near its resonant frequency and looking at the voltage across a largish feed resistor with a simple diode detector.
What shape is this thing? What response were you hoping for and why?
I presume you mean that you are hoping to excite your aerial design with the 5 or 6th harmonic of a sine wave input. This is not likely to be very rewarding unless it is an incredibly poxy sine wave source!
A square wave signal would at least give you plenty of odd harmonics.
Build a simple RF oscillator and repeat your experiment (illegally broadcasting) or perhaps legally connect a wide band RF scanner to the antenna and see where it pulls in signals from the ether (or not).
ie broadcast a weak signal at it locally and monitor reception.
The uglier the antenna, the better it works. Fractals are too nice looking to work well.
Thank you for describing everything except what you actually are trying to build and what equipment you have to work with. You're not going to be able to do anything useful with a 20MHz function generator.
Yep. Let's go back to basics. Resonance is when the leading inductive and trailing capacitive components of the antenna impedance cancel, leaving only the purely resistive component. The resistive component can be any value. As long as the inductive reactance and the capacitive reactance are equal, the antenna is deemed resonant. Now, that is not necessarily the optimum operating point. It's also probably not the point where the antenna has the best radiation efficiency, the best bandwidth, or the lowest VSWR. Details:
All the being fractal does is give you a cool looking antenna. According to the tradition fractal antenna promoters, the benefit of a fractal antenna is to cram the most gain into the smallest volume. The effectiveness of this approach is highly argumentative and has been beaten to death in rec.radio.amateur.antenna newsgroup. Search the Usenet archives for Dr Nathan Cohen.
You have nothing. You have no way to measure antenna gain, VSWR, reactance, or bandwidth. You also have no way of comparing various possible fractal antenna designs. At a minimum, you'll need an antenna analyzer.
Yes. Sell your function generator and buy an antenna analyzer, network analyzer, RF sweep generator, spectrum analyzer, directional coupler or VSWR meter, plus whatever is necessary to do a proper field test. This is old stuff, but is roughly what you need to design and test antennas: I can provide some specific equipment recommendations once you decide to disclose some numbers describing what you're trying to accomplish.
You'll also need to run a computer model of your fractal antenna. I suggest a fast PC running 4NEC2 (free). It includes some fractal antennas with the download: There are other NEC2 based antenna modeling programs available, but
4NEC2 is my favorite.
We're not done yet... You also need a reference antenna. The exact construction varies with frequency, but basically, it's an antenna with known characteristics to compare your fractal creation. If your numbers say that your fractal thing has 3dB more gain than your reference antenna, then attach a transmitter or receiver to each antenna and prove it. If that fails, then fix your design or model.
Good luck.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
While you're stocking up on all the good gear, don't overlook the benefits of doing some homework and improving your understanding of the theory involved - transmission lines, thevenin and norton networks, and some basic complex algebra. The better your understanding of what's going on, the better will be your outcome, with or without the gadgets.
Here is a reference that might help you a bit. I think your expectations of what to look for may be somewhat askew.
Easier to compare the new antenna performance against some existing one on a distant weak signal like weather reporting for an airport.
I am not convinced that you can't get quite a long way with a bit of native cunning and some luck. There are many designs on the web and elsewhere for airband antennas of various shapes and sizes. I found by accident that despite being hopelessly mismatched the right length of bell wire as a half wave dipole loop works surprisingly well for receive only. (much better than the wideband stubby whip aerial)
It makes me wonder what you can actually get away with if you are not going to transmit and only want to receive signals. I keep meaning to try wirewrap wire on the same frame to see if it will also behave OK.
The other amusement is that this crude dodgy antenna still manages to have some sensitivity to harmonic frequencies where it should in theory have essentially zero response. The theory and practice seldom quite agree when you have longish wavelengths and ground underneath.
My VHF dipole for the hifi radio tuner isn't exactly wired correctly by the book either but it still provides more than enough signal.
I beg to differ. At VHF, using a distant weak signal from an airport guarantees that something will inside the Fresnel zone and cause edge diffraction problems. At HF frequencies, ground wave propagation is highly affected by literally everything along the path. Lastly, working with weak signals is difficult. Better a strong signal that can be seen on a spectrum analyzer or frequency selective voltmeter than to be guessing whether one is measuring signal, noise, or both.
I conveniently live on the side of a hill, where there are numerous transmitters on a mountain ridge line on the other side of a valley. With the valley in between, there are few obstructions inside the Fresnel zone. The signals are stable, strong, and I can move around without seeing the usual peaks and nulls characteristic of edge diffraction. It's not perfect because of numerous trees, but it's much better than trying to do it across flat land.
When comparing similar HF antennas, it's often difficult to get a decent reading on a spectrum analyzer. There are just too many things in the way that have an effect. The best I can do is an RF field strength meter located about 300ft from my house at an elevation that hopefully puts it in the middle of the takeoff angle of my test antennas. I can see the meter with binoculars or telescope. Highest meter reading wins. In the past, I had 500ft of RG-6/u coax run from my spectrum analyzer to a remote antenna "probe", but the mice and squirrels chewed up the coax cable.
True. Any random piece of wire will radiate. The iPhone 4 antenna is a good example of that. The trick is getting the maximum performance out of the antenna, given the limitations of the package and location. Gain, size, bandwidth... pick any two.
Retch. Bell wire is steel coated and not the best material for airband (108-136MHz) antennas. However, I've made coat hanger antennas that sorta worked. Such methods work well up to about 150MHz, where cut lengths aren't particularly critical, but fail miserably at much higher frequencies.
The difference between a receive only and a tx/rx antenna is that the receive only antenna can have a very high VSWR, and still work just fine. A tx/rx antenna needs to have a fairly low VSWR in order to not kill the transmitter output stage. For example, a Beverage antenna is a receive only antenna, that does everything right, except have a low VSWR. Of course, an antenna tuner can match almost anything, but the losses tend to be rather high at lower HF frequencies.
Wrong. A 1/4 wave whip antenna is a 3/4 wave antenna at 3 times the frequency. The antenna pattern may look like a clover leaf instead of the traditional donut, but the antenna will work at higher harmonics. Many HF antenna take advantage of this effect on the harmonically related ham and marine HF bands to simplify the design.
Mediocre performance is easily achieved. Superior performance takes calculation and effort.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Even at mid-VHF (FM), the band noise is quite strong, so loosing one or two decibels is not so catastrophic. On mid-MF band (AM) even a ferrite rod with -40 dB gain is usable due to the high band noise.
Wavelength, capture area, efficiency, directivity and gain are interrelated parameters, Knowing the rest of these parameters, one missing parameter can be calculated from the others.
Unfortunately, usually "gain" and usually implicitly wavelength is known, but the Rx/Tx characteristics can't be determined from these only.
For a receiving antenna, it would be natural to specify effective capture area, so it would be easy to calculate the input power [W] for a specific power density [W/m²]. Previously, the antenna effective height [m] was used with field strength [V/m] to calculate the input voltage [V], but this had various impedance issues.
For a transmitting antenna, the gain is defined as directivity multiplied by efficiency. Since measuring directivity from the half power beam width is easy, most snake oil antenna companies simply express the directivity in decibels and claim this is "gain", completely forgetting the efficiency in the definition, which tend to be quite low in any "exotic" construction.
Regarding SWR (and reflectivity coefficients) bandwidths, it is easy to design an antenna with SWR less than 1.5:1 for DC to "daylight" with sufficient losses, a good 50 ohm dummy load is such things at least for a few GHz, as I tried to imply in my previous post :-).
Indeed, depending on antenna type, resistivity isn't as bad as it sounds. An inductor made of stainless steel can easily have a Q less than one; it's a dreadful material, well suited to resistors (nichrome is simply stainless steel, sans the steel). But an antenna which couples into space efficiently enough that it already has a Q as low, won't mind nearly as much if it's made of iron or copper. A resonant dipole, for being "resonant", is surprisingly wideband, and doesn't mind too much about the material.
IIRC, it's something like a 13dB (i.e. factor of ~20) difference in resistivity, yet only 5dB or so difference in gain, not something you want to use for transmission (unless you like heater coils for antennas!), but not terrible for recieve.
It also depends where: if you made the center 20-40% of the dipole out of copper, and the ends out of who cares what, you'll get performance almost identical to a full copper antenna. Or if you make the center portion thicker, etc.
Tim
--
Deep Friar: a very philosophical monk.
Website: http://seventransistorlabs.com
Yeah, sorta: Power efficiency and radiation efficiency are usually specified separately. For example, this junk antenna, mostly made from alternating lengths of 1/2 wave coax cable: shows 100% power efficiency but only 60% radiation efficiency thanks to the high VSWR. Note that it's not caused by the coax cable losses since NEC2 cannot correctly model coax cable. In the above antenna, the predicated gain includes the crappy radiation efficiency, which is the way it's normally specified.
Radiation efficiency is really bad with short loaded HF mobile antennas. These antennas are small fractions of a wavelength long and have big lossy inductors: See table 22.21 for how bad it can get. With extremely low radiation resistance and high current, the I^2*R losses can be rather high. Since there's no ground connection in a mobile, ground losses are high. Anyway, the calculated gains do include radiation efficiency.
Unless I'm missing something, the gain calculations include radiation efficiency losses.
What I find amusing from the snake oil antenna companies are the creative variations on radial scale of the polar antenna plot. While a log scale, with 10dB per division is fairly standard, some manufacturers will use a linear scale, which makes the pattern look very sharp and narrow. Others use idealized NEC2 or NEC4 plots, instead of actual test results.
Yep. The radiation efficiency of a dummy load is 0.0%. Lots of RF power going into the load and nothing out.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
If you have a high VSWR, there is usually (but not always) a strong reactive component and only the resistive part can dissipate power either to heat or radiate into the space).
What is the point of specifying gains for such simple vertical antennas, the efficiency is the only thing that really matter for those physically small antennas.
Those "gains" claimed in advertisements rarely include the efficiency and for these reason ARRL did not refer to those claimed gains in their articles.
Few antenna companies have a sufficient antenna range for the wavelengths involved for comparative testing with a calibrated reference dipole.
For instance measuring maritime/aeronautical beacons in the LF band might require flying circles with different radiuses at different altitudes around the antenna in a small private plane. After a series of flights it was concluded that those 90 m high masts had an efficiency about 1 %, which was considered a sufficient result.
These days a radio controlled plane might be usable for measuring MF/HF verticals.
True. But, we were talking about building antennas using lossy steel wire. With unplated steel wire, it is possible to have a purely resistive antenna, and still have I^2*R losses convert some of the RF into heat.
Gain is specified because:
gain is easy to understand.
unsophisticated customers expect a gain specification.
gain offers an easy standard for comparing antennas. More gain is more better.
gain is difficult to measure thus allowing some creativity in characterizing the antenna.
polar gain plots look cool.
specifying gain avoids having to specify embarrassing losses.
For HF antennas, you're quite correct that gain is rather meaningless. The gain of an HF monopole is fairly flat until the length drops to below about 1/10th wavelength making it easy to specify. However, the dissipative and matching losses climb really high with short HF antennas, and environmental influences (ground losses) become predominant. For such short antennas, a gain figure would more correctly reflect the method of construction and the installation environment, than anything useful about the antenna.
That's only a problem for HF antennas below about 14MHz, where the losses far exceed the antenna gain resulting in lousy efficiency. Above 14MHz, the mostly resistive losses are greatly reduced, and efficiency tends to approach 100%.
Personally, I would like to see an NEC2 deck for the antennas reviewed in QST, so I can analyze what it's really doing. I've reverse engineered a few for practice, but I'm never quite sure I've modeled it correctly. For example, I reverse engineered the MFJ-1800 2.4GHz yagi, and found that it was really a 200 ohm antenna, not properly matched, and slightly lacking in the claimed gain: That is the type of detail I would like to see in QST and QEX reviews.
They do their best with what's available. In the 1970's, we would trudge out to a cow pasture in east San Jose to run various tests. Dragging all the test equipment, generator, and beer to the site was not my idea of fun. HF measurements would change with the season, height of water table, time of day, position of the moon, etc. Favorite piece of antenna test equipment was an HP 4815A vector impedance meter. Gain tests were done by comparing with a dipole using marine radiotelephone shore stations, a remote receiver back at the factory, and a field strength meter. The FSM worked best.
Yep. Again, at low HF frequencies and with short antennas, the antenna impedance is very low. Matching that to 50 ohms results in rather high antenna currents. I^2*R losses are therefore rather high for such antennas.
We have a marine beacon transmitter in the harbor. About 80 watts out at about 300Khz into a 10 meter whip antenna as I vaguely recall. I measured the field strength and calculated backwards to the radiation efficiency. I don't remember the result but it was far less than 1%.
Much (but not all) of the certification requirements for an AM broadcast station is done using theoretical calculations. Getting consistent and accurate numbers by measurement is difficult and usually not worth the effort. I'm told that the computah models are good enough and MUCH easier.
Reminder: RF is magic.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Not needed. Fog, rust, salt, mold, mildew, green slime, and dirt on the antenna would prevent voltage buildup and corona effects. I don't recall ever seeing a clean porcelain insulator at the base. However, my standard RF tester (4 watt fluorescent tube) did light up nicely near the antenna.
The LF beacon at the Santa Cruz harbor mouth went away around 1986. I have some photos (somewhere) but they're nothing thrilling. Just a rusting metal box and a tall monopole.
I did find a 1977 photo of the former beacon transmitter at the nearby Moss Landing Harbor: This one is much more elaborate than the former Santa Cruz harbor rusting box. 25 watts, 298KHz, 85ft base loaded vertical, 10 mile range.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
It is still a reasonable guide if the antenna has any worthwhile amount of gain it will bring in weaker signals that are completely beyond the reach of the low gain antenna or are just barely intelligible.
Obviously you trade omnidirectional for a narrower beam with more gain.
A fair point. Did they ever solve the iPhone 4 problem with having a hand and a head in close proximity when it is being used?
UK bell wire covers a multitude of sins but isn't usually steel coated just plain copper twin cable with limited current carrying capacity.
The question remains in my mind as to whether the specified diamters of the conductors in published designs matter if the aerial is RX only. (obviously the inductance to self capacitance ratio goes haywire)
I am only interested in receive only. It doesn't need to handle power.
What surprised me was that it had sensitivity at twice the nominally tuned frequency! That is it was tuned for 125MHz civilian airband
118-136 but it also picked up the military on 240-260 without any difficulty. I had expected it to have a null on the second harmonic.
It is only 2dB below the maximum signal the reciever can handle. Doing it right would risk overloading the front end on local transmissions.
Incidentally, with airport antennas, the ideal antenna pattern is a flattened hemispherical pattern, where the signal has the most gain near the horizon, but is still functional for overhead aircraft. The typical torus (donut) pattern found with vertical monopoles doesn't really work well.
I don't know. After Steve Jobs declared that all phones had the problem in 2010, I ran my own test: The average detuning caused by gripping the antenna was about -8dB at
800MHz. The iPhone 4 dropped about -20dB at 1900MHz. Adding a rubber "protector" reduced this to about -14dB. Offhand, I would say it was not solved.
However, I've speculated that the unusually large drop in receive signal on the iPhone 4 was not totally due to detuning the antenna. For one thing, the antenna apparently is not tuned. It seems to be a random length of metal trim, surrounding the edge of the phone, with no obvious relationship between length and the frequencies of operation. At one point, I speculated that the receiver might be slightly regenerative and that gripping the antenna might "quench" the regenerative oscillations. However, I couldn't prove or disprove the theory because none of my friends or customers would let me dissect their iPhone 4. For me, the large drop in receive signal remains a mystery. To Apple's credit, they have re-enabled the field test mode, so that users can make their own measurements instead of having to rely on the number of bars.
Skin effect is what largely determines the resistive losses in an antenna. Clean copper plating over steel, as in copperweld, makes good antenna wire. Copper oxidizes into green or black surface crud, which makes poor antenna wire. Except for silver, most metallic oxides make poor antennas. However, there's a question of degree. In the distant past, I tried various metal objects (hand rails, rain gutters, barbed wire, water filled garden hose, and worse) for their effectiveness as HF antennas. Most worked just fine, although not quite as good as clean copper or copperweld. My guess would be that the criteria for an effective HF antenna is design, location, matching, construction, and materials, in that order.
The diameter of the antenna wire determines the bandwidth of the antenna. When low frequency broadcasts were the fashion, cage antennas were used to increase the effective diameter of the wire.
However, that doesn't mean you can build a usable antenna out of #44 AWG wire. As the wire gauge decreases, at some point, the wire resistance becomes significant, turning the antenna into a giant fuse (in transmit). It really becomes a problem with small magnetic loop antennas, where the RF currents are really large, and even the resistance of solder connections becomes significant. Instead of wire, these are often constructed from bent 1/2" copper water pipe.
No null. You have an end fed 1/2 wave monopole antenna at 2x frequency. The pattern is much the same as a center fed 1/2 wave dipole. There are a few tricks to using one, but nothing exotic: The nice part is that it doesn't really need a counterpoise or ground plane under it to work.
I meant that as a generalization. If you're trying to squeeze the last bit of theoretical performance out of an antenna system, then it has to be built from the best possible materials, designed to optimize whatever it is you're trying to accomplish, and installed in a manner that would not detune the antenna. However, if you're willing to lower your expectations, and find convenience to be more important than performance, almost any wire conductor will suffice as an antenna below about 300MHz. Above that, things become more critical and poorly designed antennas don't work at all.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Not that i am all up on fractal antennas, but isn't it the point that they do NOT have resonances? Just rather wide bandwidths?
A couple of years back there was a grad student with a thesis on fractal antennas and looking for some help on using Foster's equivalents to spice the antenna. If you can find that thread you might find out some more.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.