"Reg Edwards" wrote in message
news

"Tom Ring" wrote
Here's the one that works best for long booms -

G = 10 log (5.4075 B + 4.25) for B GT 1

Where G is gain in dBd and B is boomlength in wavelengths.

=======================================
Thanks Tom,

If we have 5 or 6 elements spaced approx 1/4-wave apart, that makes a
boom length of 1 wavelength and a gain of 10 dB relative to a dipole.

Can experienced users of Yagi design programs confirm that 10 dB is in
the right ballpark?

Yes please, I would like to see the equation for short booms and fewer
elements on your return from the salt mine.
----
Reg.


Reg,
I have run EZnec on a KLM 50-7LD. This is a 7 element on a 20 foot boom.
Almost exacty 1 wavelength long. There is a dual driven element; so, it is
more like a 6 element beam. The gain in free space computes to 11.01 dbi.

Tam/WB2TT

Ian White GM3SEK wrote:


There isn't really "an" equation for gain versus boom length, because
boom length is only an indicator of the potentially achievable gain. But
nothing "makes" a yagi deliver any particular gain figure.

Quite the opposite: very often, yagi designers will deliberately
sacrifice some forward gain to obtain better performance in other
important respects. These can include a cleaner pattern, a more
convenient feed impedance and better tolerance to element loading by
rain and ice; and all maintained over a wider bandwidth.

Yagis that have been single-mindedly optimized for maximum gain tend to
be poor performers in other respects - especially tolerance to element
loading.

The equation that Tom quoted is just the best-fit line through the gain
figures for one particular family of reasonably good long yagi designs.
Other good designs will have similar gain, but different performance in
other respects. And for every good yagi design, there are many more that
come nowhere near this level of performance.


Ian is correct, and I should have stated that this equation is for well
designed yagis. Also there were 3 families of antennas involved, plus
some odds and ends. The K1FO models all are within about 2 tenths of a
dB of the line in question.

Also, if we move to pure gain number, as in the multiplication factor,
the line produced by well designed yagis is virtually ruler straight
above 1 wavelength, and up to a a point I can't remember. It's above
10. And then it starts to roll off, which makes sense. I'd have to go
look at the models again to find where the rolloff becomes noticable.

tom
K0TAR

Reg Edwards wrote:

"Tom Ring" wrote

Here's the one that works best for long booms -

G = 10 log (5.4075 B + 4.25) for B GT 1

Where G is gain in dBd and B is boomlength in wavelengths.


=======================================
Thanks Tom,

If we have 5 or 6 elements spaced approx 1/4-wave apart, that makes a
boom length of 1 wavelength and a gain of 10 dB relative to a dipole.

Can experienced users of Yagi design programs confirm that 10 dB is in
the right ballpark?

Yes please, I would like to see the equation for short booms and fewer
elements on your return from the salt mine.
----
Reg.




This one by Bill Myers, K1GQ, is supposed to work well for short yagis.
I have no experience with it. All of mine is with V/U HF yagis.

G = 3Ln(B) + 9.85

tom
K0TAR

"Ian White GM3SEK" wrote
There isn't really "an" equation for gain versus boom length,
because
boom length is only an indicator of the potentially achievable gain.
But
nothing "makes" a yagi deliver any particular gain figure.

=======================================
Ian,

Yes, there must be such an equation. Indeed, there must be a dozen of
them.

I am aware that there are a multitude of things which are, or can be
optimised in the design of Yagis.

I am not interested in how gain can be maximised but what it has
actually turned out to be in practice over the years.

It is a statistical function of the number of elements or of boom
length - which are much the same as each other.

It is just a historical fact which all experienced designers must have
a good idea of but, it seems, are reluctant to divulge.

You probably have a good idea yourself. What is it?

For 2, 3, 4 .... N elements?

Ball park accuracy is too good for it.

I would like to include it in the notes to a practical, general
purpose, computer program which estimates received signal strength
from Tx power and ionospheric radio path distance. It could be titled
"How to use S-meters for Novices". ;o)
----
Reg.

Reg Edwards wrote:

"Ian White GM3SEK" wrote
There isn't really "an" equation for gain versus boom length,
because
boom length is only an indicator of the potentially achievable gain.
But
nothing "makes" a yagi deliver any particular gain figure.

=======================================
Ian,

Yes, there must be such an equation. Indeed, there must be a dozen of
them.


What I was trying to say was that there is not any single equation
relating gain to boom length. In the normal sense of the word
"equation", there is no equation at all.


I am aware that there are a multitude of things which are, or can be
optimised in the design of Yagis.

I am not interested in how gain can be maximised but what it has
actually turned out to be in practice over the years.

Ah, then you don't really want equations at all - you want historical
anecdotes.

Years ago, I used to collect gain data for long yagis... or at least,
such verifiable data as were available at the time... and plotted all
the dBd figures on the same piece of graph paper against log(boom
length).

The result was a scatter diagram showing a general increase of gain with
boom length. For well-designed families of long yagis (eg the DL6WU
series) it was possible to draw a good straight line; but there were
always a few yagis above that line, and many poorer designs below it.

It is a statistical function of the number of elements or of boom
length - which are much the same as each other.

It isn't "statistical" in any strict sense of that word. It's just a
collection of various isolated individual attempts to make a decent
yagi, and all the data include significant measurement errors.

It is just a historical fact which all experienced designers must have
a good idea of but, it seems, are reluctant to divulge.

Largely because it's all much less meaningful than you imagine.


You probably have a good idea yourself. What is it?

This was before the Internet and before computer modelling, so all I
have is an old piece of graph paper. I stopped adding to it in about
1985, when computer modelling became capable of far higher accuracy than
amateur measurements.

If anyone wants to try a modern correlation, there's lots of data in
the VE7BQH collection:
http://www.ifwtech.co.uk/g3sek/diy-yagi/ve7bqh.htm

VE7BQH's data are all for yagis that might be useful for VHF/UHF
moonbounce, so boom lengths start at 1.8wl and range up to 8wl. (This
turns out to be no use to Reg - see later - but it's an interesting
topic in its own right.)

If anyone is interested in doing it, use the L(WL) and GAIN(dBd)
columns. With Excel, it should take all of ten minutes to get the data
from the web page and into a "scatter X-Y" graph.

All of these modern long yagis are the results of computer optimization,
by people who already know what other designs are achieving. This tends
to make all the results bunch closely around the trend-line for the best
available performance. Unlike my old plot of stone-age yagis, very poor
designs don't make it into VE7BQH's table at all.


Back to Reg:
For 2, 3, 4 .... N elements?

Ball park accuracy is too good for it.

For short yagis, why not use the figures in N6BV's HFTA program? (ARRL
Antenna Handbook)

Antenna dBi Boom length @ 14MHz (ft)
Dipole 2.15 -
2-Ele. 5.5 8
3-Ele. 7.0 16
4-Ele. 8.5 26
5-Ele. 9.5 40
6-Ele. 11.0 60
8-Ele. 12.0 80

That is a fair set of "representative" figures.

I would like to include it in the notes to a practical, general
purpose, computer program which estimates received signal strength
from Tx power and ionospheric radio path distance. It could be titled
"How to use S-meters for Novices". ;o)

So - finally - it all comes out. If the context is "ionospheric", you're
not interested in long yagis at all. Why didn't you say so in the first
place?



--
73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

On Sun, 23 Apr 2006 09:57:09 +0100, Ian White GM3SEK
wrote:


If anyone wants to try a modern correlation, there's lots of data in
the VE7BQH collection:
http://www.ifwtech.co.uk/g3sek/diy-yagi/ve7bqh.htm

VE7BQH's data are all for yagis that might be useful for VHF/UHF
moonbounce, so boom lengths start at 1.8wl and range up to 8wl. (This
turns out to be no use to Reg - see later - but it's an interesting
topic in its own right.)

If anyone is interested in doing it, use the L(WL) and GAIN(dBd)
columns. With Excel, it should take all of ten minutes to get the data
from the web page and into a "scatter X-Y" graph.


Ian, I have plotted the data in VE7BQH's table, along with a
polynomial (1) curve fit and also the function given by K0TAR for
comparison.

The plot is at http://www.vk1od.net/lost/VE7BQHYagiGain.htm .

It would be dangerous to assume the fit function applies much below
the limits plotted.

Owen
--

Reg,

I don't think there is a 'formula' that is used to predict gain based
on boom length. Several experienced experimenters with boom antennas
have stated that the number of elements varies the gain very little
for a given length, however, given some general design, a number of
elements is typically half the gain of the same design with twice as
many elements and twice the boom length.

In other words, a 6 element beam of some given element spacing has a
(number from the air) gain of 6db. A twelve element version with
twice the boom length will be in the ballpark of 12 db gain.

Given that some 1.5 lambda boom antennas are made for wide bandwidth
and others are fine tuned for maximum gain, one cannot have an
accurate, or ballpark formula for beams that fit all.

One example is a man who was advertising that his j-pole antenna
outperformed a 3 element beam. His comparison was to a 3 element beam
designed for 180 deg beam width. It had two reflectors and one driven
element, but no director. It had a gain of 2.5 db, if I remember
correctly. However, the typical 3 element beam a ham would use is 6-8
db gain and has one reflector, one driven element and one director.
It's beam width is much narrower than the 180 degrees desired by the
man trying to broadcast his bootleg FM BC station.

I hope this helps.

Let me know.

73
--
73 for now
Buck
N4PGW

On Sun, 23 Apr 2006 08:08:46 -0400, Buck wrote:

[snip]

In other words, a 6 element beam of some given element spacing has a
(number from the air) gain of 6db. A twelve element version with
twice the boom length will be in the ballpark of 12 db gain.

You want to rethink this? I've always figured 2.5 - 2.6 dB increase
for each doubling of the boom length (or array size) for practical
purposes.

On Sun, 23 Apr 2006 06:17:35 -0700, Wes Stewart
wrote:

On Sun, 23 Apr 2006 08:08:46 -0400, Buck wrote:

[snip]

In other words, a 6 element beam of some given element spacing has a
(number from the air) gain of 6db. A twelve element version with
twice the boom length will be in the ballpark of 12 db gain.

You want to rethink this? I've always figured 2.5 - 2.6 dB increase
for each doubling of the boom length (or array size) for practical
purposes.

That's pretty close. I don't know the exact numbers and my numbers
are just ballpark. yours may very well be more accurate. I haven't
tried modeling for accuracy, just what I remember reading about
antennas. If you build a 2 lambda antenna and a 4 lambda antenna, they
the 4 will have about the same gain as a pair of 2 lambda antennas
phased, which is about, but not necessarily exactly double the gain of
one antenna.

give or take 1/2 db, what is that to signal strength anyway?

thanks

Buck


--
73 for now
Buck
N4PGW

"Buck" wrote

give or take 1/2 db, what is that to signal strength anyway?

=========================================

It's about 1/12th of an S-unit on your S-meter. Just the thickness of
the meter needle.

It matters even less when plus or minus 2 or 3 S-units, or 15 dB, of
fading occurs on radio paths through the ionosphere. ;o)

Why all the worry about antenna gains within about 0.01 dB ?

Just get some wire high up in the air and enjoy yourself.
----
Reg.