# antennas

Discussion in 'Electronic Design' started by RichD, Oct 2, 2007.

1. ### RichDGuest

When I look at consumer wireless devices, I ask
myself: does any actual 'design' go into the antenna?

Look at the antennas, they are just a length of wire,
or sometimes a loop of wire. I recall my days
studying EM theory, and RF design, it's complex
stuff. (though I had only rudimentary exposure to
antenna design) Then I observe what's in use, and
I scratch my head. Are there any formulae, principles,
optimizations, involved in these devices?

If you look at the hobbyist literature, you see
much the same: a circuit diagram, with that little
'antenna' thigamajig. It's just given that way, with
no information. Like manna from heaven, or what?

In aerospace, with their big \$\$, of course they do
sophisticated antenna development. But what

2. ### Jeff LiebermannGuest

Oh yes. Considerable design. With antennas, it's called "antenna
modeling" using NEC2 or NEC4. See shopping list of suitable programs
at:
<http://www.cebik.com/model/nec.html>

If you want to play, I suggest 4NEC2, which comes with a large
collection of sample antennas:
<http://home.ict.nl/~arivoors/>

Some of my tinkering with 4NEC2 and antennas:
<http://802.11junk.com/jeffl/antennas/
Or radiating surfaces. However, their simplicity is deceiving. It's
very easy to design a marginal and inefficient radiator, one that is
optimized within the required constraints is far more difficult.
Naw... All RF is magic.
Oh yes. Even the simple antennas follow the rules. The basics are
simple. You're responsible for delivering RF from perhaps a 50 ohm
source and matching it to the impedance of free space (377 ohms), in
the desired direction, and without losing any in the process.
I don't see a problem. What antenna information were you expecting on
the schematic diagram?
Your cell phone antenna system is the nightmare of antenna design. My
phone PDA has internal antennas for 900/1800MHz, 2.4GHz Wi-Fi, 2.4GHz
Bluetooth, and 1.5GHz GPS. Some new handsets will be arriving with
2.5GHz Wi-Max. All that works simultaneously, without mutual
interference, maintaining FCC Part 15 radiation requirements, SAR
(specific absorption rate) safety requirements, while delivering the
bulk of the RF in the general direction of the cell site. Oh, I
forgot that it also has to work with varying body, hand, head, and car
kit capacitance and blockage. All this within an insanely small
package. The final result may look simplistic, but are invariably the
end result of substantial modeling, cursing, and compromise.

3. ### mpmGuest

Absolutely!
For any receiver antenna design, you have to take care of opertating
frequency and bandwidth (even multiband operation), then "Q", then
efficiency issues - such as anticipted detuning if worn on the body
(i.e. Walkman type radio, etc..) It's not as simple as it might
appear at first.

A colleague of mine just finished work on a "simple" loop antenna for
a major manufacturer of new portable HD Radio receivers. So you can
conclude that even "simple" antenna design work is still in demand....
I recommend:
Weeks, W.L. "Antenna Engineering" New York, McGraw Hill, 1968 among
others as a resonable start. (non-hobbyist)

-mpm

4. ### Rich GriseGuest

In a broad sense, anything that gets built had to have been designed.

Try "The ARRL Antenna Book" - there's probably one at a local library,
or you could probably order one.

Good Luck!
Rich

5. ### Jeff LiebermannGuest

Choke, cough, sputter, etc... I need a rant. I've done cleanup jobs
on products that never saw the benefits of a calculation. Many were
reverse engineered or cloned, with only a minimum understanding of the
original design[1]. The rush to market has created some truly amazing
implementations that border on butchery. In the broad sense, I agree
that most things eventually are designed, calculated, re-designed,
re-calculated, optimized, cost reduced, cost reduced some more,
butchered, and delivered. Many products are a basically good idea,
badly implemented, and held together by a mess of band aids. Touch
anything, and the house of cards falls over. Fortunately, it was
fixing such butchery that kept me in business for a long time.

My business card still reads "If this stuff worked, you wouldn't need
me".

Since we're talking about antennas, has anyone ever tried to model the
typical Radio Shack yagi TV antenna? One would expect that with such
a long product lifetime, years of evolutionary development, the
availability of sophisticated modeling tools, and the need for more
sensitivity with OTA digital TV, that such antennas would be a
superior design. No so. I've reverse engineered two of these (that I
have on my roof) and found them to be abysmal. On some channels,
there's more gain in the reverse direction than in the forward. The
herringbone log periodic attempt is really quite omni directional with
the added detriment of lousy gain. From my numbers, it would seem
that these are designed on the basis of aesthetics and manufacturing
costs (using identical length elements) rather than performance.

[1] Incidentally, what is the one thing that you can't determine by
reverse engineering a product? Answer: design and production
tolerances. That's where things usually go awry and why industrial
espionage is still popular.

6. ### CptDondoGuest

<http://www.enterprisenetworkingplanet.com/netos/article.php/3696296>

7. ### Jeff LiebermannGuest

Cute, true, and getting worse. However, the above article is about
software abominations, which is quite different from hardware and
antennas. Software can be fixed, which is it's own punishment as the
fixes are arriving much faster than the products. There isn't
anything I've bought with software or firmware, that didn't require an
update on arrival.

Not so with hardware. They're not called patches, updates, or fixes.
They're called recalls, rebuilds, or warranty replacements, with all
the detrimental implications. Software is easy to fix, but not
hardware.

So, one would expect that there would be more care applied to hardware
design, to avoid the warranty recall experience. That's generally
true, but there were enough exceptions to have kept me in business.
The generally higher level of quality is probably due to the minor
fact that anyone can see if the hardware is malfunctioning. However,
software bugs and oddities are not so easy to see or identify.
Hardware also requires considerable time and effort to add features,
while in software, feature bloat is epidemic. Features and functions
get added faster than bugs get fixed, so the inevitable result is a
bloated and bug infested product. If that were true for hardware,
civilization would have collapsed long ago.

Antenna design has other important advantages. You can't see it work.
You can't tell how it works. Product comparisons are almost
impossible. Nobody understands the numbers. RF and antennas are
indistinguishable from magic. I've often considered going into the
antenna business, where the basic plan would be to design maximally
weird looking antennas, with marginal performance. If I have time,
maybe I'll do some calculations to see how badly it works.

8. ### CptDondoGuest

You also forgot the advertising hyperbole and inflated prices.

But in any case, I suspect Terk has you beat on all counts.

9. ### Jeff LiebermannGuest

Sorry. I didn't want to reveal the entire business plan. Yes, there
will be the requisite inflated claims, contrived testing, invented
techy-like terminology, irrelevant testimonials, and insane prices
(shipping and handling extra). I'll also try to make sure that my
antennas self deteriorate, include an unfathomable warranty, and are
delivered from a factory in China that can't easily be located or
sued.
Nope. They're dull and ordinary products. My antennas will in colors
sufficiently ostentatious to attract attention, so that everyone knows
that the owner overpaid for the antenna. I may even throw in a manual
something based on the \$1,200 AC power cords might be profitable:
When the market is saturated, I'll recycle the business plan on
midnight television. Get rich replacing all your friends and
neighbors antennas.

10. ### mpmGuest

You know what comes to mind immediately..?

Those stupid little stick-on "antennas" that are supposed to boost
All you do is take out the battery and stick this metalized sticker
underneath it to "enhance" the phone's performance. I mean, this baby
doesn't even have wires!!

And of course, they do absolutely nothing.

Maybe the business plan here screams designing stickers in the shape
of "rabbit ears" for new video cell phone services such as V-Cast,
MediaFLO and the like...??
-mpm

11. ### Jeff LiebermannGuest

Ummm, no. My crystal ball is at the sorcerers being repaired. My
abilities to read minds and predict the future are temporarily
limited.
RF is magic. It doesn't matter if it works or not. It just "feels"
or "looks" like it's working. Perception is everything.
I don't know about anyone actually paying for one, but every cell
phone battery I purchase from Hong Kong, seems to get shipped with
one. My guess is they have a surplus and are just unloading them.
Don't forget the little wire screen stickons that go over the earpiece
and allegedly protect against RF induced brain rot. Judging by the
average persons inability to operate a vehicle while yacking on a cell
phone, these don't work either.

It's interesting to note that my first adventure into radio was when I
Electronics that proclaimed "Turn your house wiring into a giant
1000ft TV antenna". Seemed reasonable to me, so I arranged to obtain
one to play with. That resulted in a nasty electrical shock as the TV
was of the AC/DC type (no xformer), and the box consisted solely of a
"capacitator". Despite my rather violent introduction to electricity,
I assumed that anything that powerful was worthy of study. It's been
downhill ever since.

12. ### Martin BrownGuest

And sometimes it is the final daft last minute cost reductions that
contribute to the most annoying product defects and excessive infant
mortality. Don't you just love beancounters...
Time to market and first mover advantage is viewed as far more
important than stable or fully functional software. Excel 2007 is a
fabulous example right now. It can barely chart a few thousand points
without keeling over or grinding to a standstill. Its predecessor 2003
is happy with charts containing tens of thousands of points and is an
order of magnitude faster for complex graphs. But so long as consumers
accept having defective new products shoved down their throats that is
what will happen.

There is an interesting corollary to the software game. Unlike
hardware, software does not wear out with age. Your requirements may
change until it no longer does the required job, but so long as it
does what you need you are much better off with an established
trustworthy application than the newest gee whiz flash bang premature
release.
Same for software development. People get into very strange scrapes
that need external skilled resource to sort out very late in the day.
Some are. I have met a few antenna designers who really know what they
are doing. Large phased arrays have to be exactly right or they don't
perform anything like to specification.

Regards,
Martin Brown

13. ### Gary TaitGuest

More or less, it depends on the device and frequency used.

Oftentimes, for low/micro power unlicensed VHF/UHF, certain basic
practice is used for antenna design, with little theory.
Depending, the antenna is specified in manufacture of the device, or as
a 3rd party component, presumably pre-engineered for the task.

Cell phones use engineered antennae, due to their frequency, power, and

14. ### Jeff LiebermannGuest

There's much more to the problem than just rush to market. Product
life cycles have decreased sufficiently that often several generations
of future replacement products are being designed when a product hits
the market. The incentive to fix any defects in the current product
is lost if someone decides that "we'll just give them the new version,
which will be out next month". Why put time and money into fixing a
product that obsolete on introduction? (I've seen this happen several
times).

Note: I am not a programmist.
Nope. Bad example. Excel is a classic example of my software axiom
"Functions and features get added faster than bugs get fixed". The
inevitable result is a bloated monster, burdened by useless features,
and full of bugs. I haven't actually tried Excel 2007, but if it's as
bad as you suggest, I'm sure it qualifies.

The effect is understandable. Functions and features sell products.
Bug fixes do not. I haven't seen any retail software package display
on the box "Fewer bugs than previous version" as a product feature.
Well, actually MS did that when Windoze 2000 was introduced and
proudly proclaimed that "it crashes less often". Anyway, if you have
limited time and resource, and you have to choose whether to allocate
them to new features or old bugs, the choice is obvious. Bugs can
always be fixed after product release.
Mediocrity sells. What I find fascinating is that buyers often cite
the vain hope that the new version may have fixed some of the bugs as
However, when I find something I like, I stay with it. I'm still
running Office 2000 and not having any major issues as compared to
those running later versions.
Ah, but there is also "software rot":
<http://en.wikipedia.org/wiki/Software_rot>
Old software just looks old. The surplus stores are full of perfectly
adequate, shrink wrapped software, that would totally functional were
it kept up to date with bug fixes. However, that's rarely the case,
so toss the old bugs, and replace them with new bugs. It's so bad,
that at least one company had released new product disguised as an
older product in order to convince consumers that perhaps the long
history might yield fewer bugs. Free Lotus Symphony:
Agreed. I'm a mediocre antenna designer and have great respect for
those that understand the technology. I read some of the IEEE Antenna
and Propagation Proceedings and am often lost in the math and models.
However, I was referring to the average consumer of antennas. TV
antennas are again a good example. Look at the variety offered and
try to distinguish by either observation or specification, which are
the best antennas for OTA TV reception. Most consumers can't do that,
so they purchase the biggest, weirdest looking, and most gold plated
yagi, that's really a fairly rotten antenna compared to a simple bow
tie array, backed by a large barbeque grill. Same with the rabbit
ears, with the tiny parabolic dish in between, that sits on the TV.
The dish does absolutely nothing and is there strictly for looks.

The problem is that sometimes, such weirdness goes too far. I once
designed (on paper) a TV antenna consisting of a helical spiral array.
Lots of gain, lots of bandwidth, polarization insensitive, but really,
really, really big and ugly. It looked like a giant conical ziggurat
on its side with a giant barbeque grill base. I didn't have to build
one as it was obvious that nobody would want one on their roof. Ugly
and fancy sell, but there's a limit.

15. ### JoopGuest

Also the hours of standby time / airtime is influenced heavily by the
efficiency of the antenna. So unless marketing dares to lie, it pays
to put some attention to the engineering of the antenna.

Personally I had a pleasant surprise lately. I bought a miniVNA
(http://www.wimo.com/instrumentation_e.html#minivna).
After it arrived I started to hook up all kind of stuff in my house.
Including the rubber duck antenna of my HAM Standard porto. Initially
I thought it was way off in frequency.

Until I put it straight up and held the miniVNA in my hand as if it
was a portofone. Suddenly it was exactly on frequency (obtimum in the
middle of the band) and very close to 50 Ohm. Wow. This thing must
have been engineered, tested and tuned.

Joop

16. ### Frithiof Andreas JensenGuest

Yes so; My dishwasher made a lot of pretty colours with a circuit board
carrying 380V so I went to the spare parts shoppe to get a New One.

What I got handed was a fixpack: A cardboard box with instructions,
replacement hoses, cable ties, a small injection pump and some rubber
gaskets to be fittes in specified locations to stop water from dripping onto
the board and make colours. The new board was coated too with some red-ish
thick varnish ;-)

17. ### Ken FowlerGuest

The description and price on the MiniVNA are very attractive. Have you had any unexpected
results or problems? Are there any limitations on R and X values, especially for HF antennas?
Have you noticed any interference from broadcast signals? Thanks.

Ken Fowler

18. ### JoopGuest

No problems. The only thing is that at low levels there can be some
small spikes traveling on top of the curves. Keeping it further from
my PC improves that. Perhaps due to the plastic housing.

Further it does not measure full 360 degree phase, but 180. This is
because an AD8302 is used. But then I find absolute phase reading not
Perhaps when the software might progress and tries to compensate it
might be more important.

I am not sure on what you mean with limitations on R/X. Anything way
off 50 ohm gets inaccurate. Whether this is worse than other simple
VNA's I cannot judge. In the Yahoo group
http://groups.yahoo.com/group/analyzer_iw3hev/ this pops up every now
and then. Different software versions seems to behave a bit different
in this respect as well. I am simply hoping it evolves in the right
direction over time.
I did help a friend with his antenna though. He had some difficulty
finding the right settings of his antenna tuner. Partly because his
homebrew PA had a tendency to oscillate. We hooked up the miniVNA an
within a few minutes we found optimum settings for all bands of
interest. Minor adjustments were needed with the PA hooked up (we know
the PA is not 50 ohm itself).

With my type of measurements I did not experience broadcast issues.
But then it is not as sensitive as e.g. a spectrum analyzer. Also my
antenna tests were not in broadcast frequency range (or no strong
signals nearby).
Do not think of it as the perfect network analyzer. Having said that
it suits me fine. Measuring crystals (Q, Rs, frequency), cable length
/ velocity factor, resonance points, bandpass curves, return loss
(range about 0-40dB), having a 1-180 MHz stable signal generator and
such make it a very nice device.
The software can export data to spreadsheet or a smith chart tool
written by another party. Screenshots can be saved for documentation
etc. The only thing is you need a computer to use it (laptop will do).
A simple PDA program does exist, but I have not yet seen screenshots.
Else it might tempt me to buy a PDA.

I guess an advantage is that the software improves over time.
Calibration (output/sensitivity curves and DDS frequency) is made a
lot easier in the latest version. But still I use several versions
depending on what I want to do. Installation can be a bit dodgy as
well, but usually people succeed and users are happy. Note I only use
the windows software. The linux version seems to be made by more
skilled software developers, but I have not used it. So I do not know
if it has similar calibration features as the windows stuff.

Joop

19. ### Ken FowlerGuest

Joop,

Thanks for the assessment of the MiniVNA. My reason for inquiring about R and X accuracy is because
I want to measure impedance versus frequency at the antenna terminals and use the R and X values to
calculate the necessary network C and L for low VSWR. Broadcast station rejection is important when

measuring out of doors at HF.

My dream is to afford the AIM-4170. See: http://www.w5big.com/ \$500 is not in my budget right
now.

Thanks.

Ken

20. ### LVMarcGuest

The design does start with the frequency and the speed of light fro
there beyond consumer goods (whom use the shit antennas they can get for
..10 cents) there is design. controlling the beam pointing and providing
and efficent conversion of EM energy t a voltage you can use to process.

And there is still room for innovation!

go to

www.fwt.niat.net

here the typical frequency and size rule of thumb...is broken and 1/8
sized volume is now possible, ad with gain and controlled impedance and
beam charactetisitcs, too!

Marc