Michael Tope" <W4EF@dellroy.com
Fri, 9 Feb 2001 15:02:50 -0800
Yuri - please see my comments below.
Thanks - Mike, W4EF.............
----- Original Message -----
Sent: Thursday, February 08, 2001 2:56 PM
Subject: Re: [TowerTalk] quads
> Hi Tom,
> looks like here we go again. I express some of my knowledge and information
> based mostly on practical experience, as response to some claims otherwise,
> and I get arguments that it can't be so because books say (and Tom knows)
> otherwise. We went through similar exercises before on other topics. The
> result usually is that Tom, gets quiet, maybe goes, does some work. Then
> comes back after a while and becomes "guru" on stuff that I was accused to be
> promoter of "pathological science" and not given credit.
> I will do this one more response to your response. I do not appreciate being
> put into the same bag with Fractenna Freak, I don't think I deserve that (nor
> the others I mentioned).
> In a message dated 2/7/2001 18:47:10 Eastern Standard Time,
> firstname.lastname@example.org writes:
> > Hi Yuri,
> > > It appears that some of the "new wisdom" is coming from the
> > > software modeling "facts."
> > Actually it comes from simple math that appears
> > in antenna engineering books from the 40's and
> > earlier.
> > A quad is two 1/4 wl end-loaded dipoles stacked
> > 1/4 wl apart.
> > Stacking gain for two elements spaced 1/4 wl is
> > one dB under **perfect** conditions.
> That means, from the times before Quad and based on the math formulas, it
> can't be? Just like Bumble Bee cannot fly? Or like hams cannot communicate on
> "useless" shortwaves?
> > > I just quickly modeled single Quad 1/4 wave/side loop vs. dipole
> > > and I can get paper advantage of over 2 dB. So no folklore there.
> > I cannot, no matter how I model it. I see one
> > dB or less maximum gain over a dipole at the
> > same mean height.
You should both publish your models for peer review (oops that too much like
> I guess that was the difference between Tesla and Edison. One could, one
> couldn't. I can get in EZNEC close to 2 dB difference.
Yuri, by this statement I gather that you have confidence in computer models
based on mathematical formulas (you can't have it both ways). If that is the
case, you should be able to model your razor, the 6 element KLM, and the 5
element Telerex that you speak about and demostrate why the razor has a 10dB
advantage over the KLM, and a 20 dB advantage of the Telerex. This could be
done without disclosing any of the details of your Razor design (you could
publish pattern data only).
> > In freespace, it is about one dB on the button.
> > At some heights, it is less than 1 dB...and
> > sometimes approaches zero dB. With more
> > elements, the advantage is less at any height.
> > Like any antenna, a quad "gets gain" by forcing
> > a null in an area of significant radiation. That
> > area is directly in line with the plane of the
> > current maximas.
> > If you place the quad at a height where ground
> > reflection forces a null straight up, the
> > horizontally polarized quad element provides
> > minimal, if any, gain.
> > If you place it at a height where radiation is
> > maximum straight up, you get maximum advantage
> > but at all practical heights that advantage is
> > less than one dB for the two antennas compared
> > at the same mean height.
> > If the model shows more gain than theory
> > (available since antennas began to be
> > understood) predicts, the model is flawed or the
> > comparison is unequal for mean height. Broadside
> > array gain has been defined for many many years
> > and the results have never been proven wrong to
> > this day.
> Well, then the model or software is flawed. I will verify that on real life
> models on 2m and 10m. I am waiting for snow to melt.
Yuri, you just made a claim based upon the results of EZNEC. I believe
this is the program that Tom uses.
> > >(My 3 el. Quad beat 7 el. KLM Log Yagi on almost twice
> > the boom on 2m.)
> > > Quad (circular or delta) loops have larger aperture than dipole,
> > > they have fatter vertical lobe and there is
> > > suspicion that they pick up some of the "other" polarization too.
> > I hear "aperture" used all the time to explain
> > why one antenna is better than another.
> > Aperture has NOTHING to do with physical size of
> > an antenna. Aperture relates only to gain and
> > frequency, not physical size.
> Here you are contradicting yourself. You state that quad loop can have, lets
> suppose only 1 dB gain, which means more gain. More gain, as you say means
> larger aperture. Then you say, it can't have it. Which is then wrong? (I
> didn't say that loops have larger apperture because they are bigger!)
The term aperture generally only has much meaning when discussing reflector
antennas or horns where the gain of the antenna relates to the surface integral
of a field distributed coherently across a radiating surface. In this case,
doubling the surface area or aperture, will nearly double its gain (assuming a similar
field distribution is maintained). For wire antennas, the analogy breaks down. In this
case you can calculate the "effective aperature"of the antenna in (meters^2), from the
gain of the antenna, but this number doesn't bear much resembleance to the size of the
physical radiator. In the case of wire antennas, the gain relates more to a line integral
than a surface integral (in this case the far field is the sum total of the field
contributions from each element of linear current flowing in the wire).
> > A two inch square antenna on 40 meters can have
> > more "aperture" than a two mile square antenna.
> > A fatter vertical lobe, for a given azimuthal
> > beamwidth, means less gain (and more fading when
> > multipath is involved).
> > If any antenna picks up two polarizations, gain
> > is always less than an antenna with the same
> > pattern width that picks up only one
> > polarization. If the antenna picks up both
> > polarizations on a skywave path, or truly
> > transmits both polarizations, fading is
> > increased...not decreased.
> > Quads do not pick up both polarizations because
> > they are not constructed in a way that lets them
> > to do that. Both polarizations
> > are received (or transmitted) only when the
> > antenna has dual elements or dual radiating
> > areas with suitable phasing and
> > polarization, like crossed dipoles in phase
> > quadrature or a small loop and a small dipole on
> > the axis of the loop.
> > Quads pick up one polarization only, although
> > that polarization might be skewed by feedpoint
> > placement or feedpoint errors. Even
> > if it is skewed, it is one polarization. It is
> > just tilted.
> Can you tell me what polarization has a quad loop that is closed, without
> feedline attached? There might be a hint.
No, but I can reduce any incident polarization into vertical and horizontal
components and linearly independent current distributions.
> One experiment I made was; I wanted to see what gain if any has a quad loop,
> horizontally polarized (fed at the bottom), diamond shape, in relation to the
> vertical tower that would go near/through it. At about 1/4 wave from the
> vertical tower, the loop exhibited about 5 dB F/B. Can the math calculate
Did you use a balun to feed the antenna? If you had significant RF current
flowing on the outside of the feedline, this could account for the front-to-
back you saw. How else would you explain this? If a quad loop with physical
symmetry about the vertical axis is exited by a vertically polarized wave, why
would current flow into one wire and out of the other since they both see
the same excitation? Which side is the up wire and which is the down wire?
> > > Quad loop has lower Q, wider frequency response, flatter SWR
> > > curve, wider bandwidth.
> > So do dummy loads.
> That is very professional response. So the "dummy" quad has gain over dipole
> (you admit that) and it is a dummy load? Can you 'splain that?
> > > Quad loops provide more efficient match and RF transfer between
> > > coax and antennas.
> > That applies to UHF and higher where the coax
> > diameter is large compared to the wavelength ,
> > but not at VHF and below.
> What has diameter of coax to do with match? More efficient match is due to
> the fact that loops have a higher impedance. When we make an array out of
> quad elements with higher impedance than in Yagis, less "transformer" is
> needed, fewer losses. I managed design my antennas to have 50 ohm feedpoint,
> no matching devices (which always have losses), better power transfer.
> Another benefit is good soldered connections and solid copper wire loops,
> rather than corroded junctions and "matches" as in aluminum Yagis.
> > > http://members.aol.com/ve3bmv/Razors.htm Single Razor would
> > > consistently beat 6 el. KLM by about 10 dB in real life tests.
Yes, that seems like a good argument to me, but the difference in feed losses
at HF between a monoband quad and monoband yagi couln't amount to more
than a few tenths of a dB, otherwise the yagi feeds would be burning up at
1500 watts [imagine dumping 300 watts (1dB) of heat into the typical gamma
> > Are you really saying you "razor" has a measured
> > gain of ten dB over a comparable size yagi?
> I didn't say measured to 0.00001 dB accuracy. I said in real life tests. That
> means real antennas, with real stations, with real (inaccurate) S-meters, but
> being able to detect differences. Averaged over period of time, numerous
> tests, various directions,
> that was the rough result. BTW KLM 6 el. Big Stick showed also about 10 dB
> advantage over 5 el. Telrex. This is what it means, real life, not with
> scientists in white coats. This was the final verification of performance in
> real life with all those imperfections that take part.
Where the antennas at the same height? At the local club station, I often
see 10 dB (somtimes 20dB) difference between a 5 element 20 monobander
and a KT34XA, but the antennas are at different heights. When the band
is long (late at night to JA), the KT34XA at 114' compares favorably with the
5 element mono at 70'. During the day to Europe where our upward sloping
terrain probably favors the lower antenna, the monobander is almost always
better by an uncalibrated S-unit or two. This is all terrain effect stuff. If I
had two really big anechoic chambers to compare the antennas with, I bet the
monobander isn't much better than 1 to 2 dB better than the KT34XA (48
versus 32 foot boom) in free space forward gain.
> > Either something is wrong with the reference
> > antenna or the test was seriously flawed. 10 dB
> > gain over a 5 element yagi would require an
> > enormous array!!!!!
> I never said that, see above.
> You have seen reports from 160m contest, and those reports have some meaning,
> you are heard, you are worked, or nothing.
> > > Some notable contesters use(d) loop
> > antennas to cream competition
> > > (W2PV,
> > > K6UA, VE3BMV, K3ZO, KC1XX, VK3MO) and their
> > experience confirms the
> > > benefits of loop antennas. Those who know ..
If loops are so great, then why did W2PV uses yagis on 40 - 10?
Same goes for KC1XX. How come K3ZO uses a yagi on 80 meters
instead of a quad. Also, W2PV stated in his articles exactly
the same description of the quad that W8JI gave - namely two
bent dipoles stacked 1/4 wave apart. He further pointed out
that for this reason, the advantage of the quad over a yagi
diminished as the boom length and hence number of elements
> > This sounds like the "science" used in the
> > infamous "Fractal Run" threads of RRAA! "I
> > worked "X" stations in "X" minutes so my antenna
> > must work better than anything else"!
> This is really low blow to all those who contributed so much to amateur
> radio. Their results are immortalized in the contest results and records.
> K6UA, K3JH had rotatable Quads on 80m, they had reports from DX and Americans
> serving overseas, that their signals were more consistent and dominating at
> varying conditions, where 3-4 el. Yagis didn't hack it. This is no "Fractal
K6UA uses a wire yagi nowadays on 80/75. When I listen to the big guns working
longpath at our sunrise here on the West Coast, the biggest standouts are W6RJ
and W6ANR. The guys carry on Q5 SSB qsos with guys nobody else can hear.
They are both running 3 element rotary yagis. Does that prove a yagi is better than
a quad? No probably not, but it goes to show that for every anecdotal account of
quad superiority, I can probably counter with anectdotal evidence of yagi superiority.
Anyway, I don't think Tom was comparing all those fellows to N1IR. He was simply
stating that anectdotal accounts don't prove that one antenna is instrinsically superior
to another. That requires "scientific" tests (ooh theres that dirty word again) which
properly control for terrain effects and propagation. I haven't followed that fractal
thing too closely, but from what I have seen, there were a lot of people saying hey
this antenna works pretty good, but no data (real measurements) showing that it
was better really better than other design of comparable physical size.
> > Anyone who deals with statistics knows such data
> > is meaningless, there are far too many variables
> > other than gain involved in the results.
> Not exactly, some meaningful data are obvious from contest results,
> especially when comparing stations in close proximity, with other factors
> being close, the antenna shows prominently, it is the tool that allows to
> extend the propagation/conditions and allows working another layer of station
> that others can't even hear. When I see my scores jump by 40% and I nail all
> Canadian Records, I know I have better antenna than 1 dB. I don't care for
> "scientific" analysis.
There is no question that your razors were good antennas. I remember hearing
your loud backscatter signal when I was in Ohio in the mid-80 "Victor Echo 3
Bravo Mike Victor", but how do you know your improving scores weren't the
result of a combination of factors - incremental improvement in antennas (razor
beams), improving operating skills, improvements in equipment (better filters),
and better conditions from year to year?
> > > I am just now "investigating" loops with
> > vertical polarization for
> > > the
> > > salty beaches, see what comes out. I am
> > already getting 2 dB from
> > > single loop over "mighty" dipole.
> > I modelled a vertical dipole over perfect
> > ground. The gain for low dipole heights was 7.3
> > dBi.
> > I did nothing but convert to a vertically
> > polarized quad at the exact same mean height,
> > and gain was 7.83 dBi.
> > In this case, where everything is equal, the
> > quad supplied a "roaring" .53 dB gain advantage.
> > I placed the antennas at the same height over
> > lossy ground, and they were equal!!!
> Maybe the same difference as between Tesla and Edison :-)
> For the situation of interest to me, I got 5.66 dBi from dipole and about 7.4
> dBi from the loop in real life situation (not in the free space that I can't
> get into for any money :-). That is closer to 2 dB than .53.
Okay so its 2dB and not 1 for a single element. Then how do you explain the
approximate 10 dB advantage of a razor over the KLM? and the 10 dB
advantage of the KLM over the Telerex when the boom lengths are
approximately equivalent (or were they?). You really have to screw up a
monoband yagi bad to be off by 10 dB (especially since the KLM probably
only has about 8 to 10dB gain at most over a dipole). I am not doubting
your observations, only whether or not you properly controlled for all
> > 73, Tom W8JI
> > email@example.com
> You know your stuff, I know (and saw with my eyes) my stuff. I enjoy some of
> your stuff, but when I see something that doesn't jive with "my stuff" I try
> to point it out (I am not alone as I saw), and then I get this...
> Usually things are discovered and then "scientists" move in and figure it out
> and explain it. So stand by...
> 73, Yuri, K3BU
> FAQ on WWW: http://www.contesting.com/FAQ/towertalk
> Submissions: firstname.lastname@example.org
> Administrative requests: towertalk-REQUEST@contesting.com
> Problems: email@example.com
FAQ on WWW: http://www.contesting.com/FAQ/towertalk
Administrative requests: towertalk-REQUEST@contesting.com