I'm getting an initial impression that there might be three basic forms of
traditional yagi (not log-periodic) when considering the issue of
bandwidth.

Common Yagi (whatever that may be) . This has a fairly narrow bandwidth,
not sure of the typical bandwidth in precentage terms of the designed
frequency. Perhaps 1% at a guess. Here there has been no effort to
concentrate on making any form of wideband antenna.

Optimised Wideband Antenna (OWA). This has a fairly wide bandwidth, not sure
of the typical bandwith maybe 3 or 4 percent of the design frequency. Here
some effort has been put into making something of a wideband antenna, but
the design is not as wideband as is possible.

Wideband Yagi. This is typically the antenna having the widest bandwith.
Not sure of the typical bandwidth, but perhaps about 7% of the design
frequency.

If you were to draw up a list of traditional yagi' designs (including call
letters of the designer if there is one) and place them in order of
increasing bandwidth, what kind of list would you get?

Here is my measly effort:

Yagi (under no classification but yagi)
OWA
Wideband Yagi.

I'm not sure where in a list you put DL6WU, W4RNL, DJ9BV, + other designs I
know not of. Does a DL6WU yagi come under OWA or wideband yagi? Has anyone
drawn up a definitive classification?

"Richard" wrote in message
...

I'm not sure where in a list you put DL6WU, W4RNL, DJ9BV, + other designs
I know not of. Does a DL6WU yagi come under OWA or wideband yagi?
Has anyone drawn up a definitive classification?

Further:

Is there in existence or has anyone ever thought or is there any possible
usefulness in a rating system for yagi antennas? You ascribe a value to
each of the main the characteristics and you publish the rating along with
the design.

Something like:

Gain range (dB): 0-2 2-4 4-6.....infinite
Rating: 1 2 3 10

F/B range (dB): 0-5 5-7 7-10....infinite
Rating: 1 2 3 10

Bandwith range %: 0-0.5 0.5-1 1-2 2-4.....infinite
Rating 10

Then you would publish in the article for design something like:

G8VOQ "Super-duper-all-singing-and-dancing yagi":

Parameter Rating
Gain: 10
F/B: 10
Bandwidth: 10

Is there in existence or has anyone ever thought or is there any possible
usefulness in a rating system for yagi antennas? You ascribe a value to
each of the main the characteristics and you publish the rating along with
the design.

Virtually all the latest Yagi-Uda work is aided by simulation and computer
optimization. This is assessed through a 'cost function' (also known as figure
of merit; objective function) which is a mathematical formula that weights
bandwidth; gain; feed impedance; F/B and so on.

There is no 'one' cost function, and tailoring it requires some finesse. The
tradeoff usually occurs in bandwidth compared to gain and F/B. The bigger the
value of the cost function, the better the candidate design.

A simple expression that could be used is

CF= aG +bF +cW (1)

where G is gain in db; F is F/B in dB; and W is bandwidth in percent. the
a,b,c coefficients thus determine which is a bit more important to you, as
well as allowing relative normalization.

73,
Chip N1IR

"Fractenna" wrote in message
...
Is there in existence or has anyone ever thought or is there any possible
usefulness in a rating system for yagi antennas? You ascribe a value to
each of the main the characteristics and you publish the rating along
with
the design.

Virtually all the latest Yagi-Uda work is aided by simulation and computer
optimization. This is assessed through a 'cost function' (also known as
figure
of merit; objective function) which is a mathematical formula that weights
bandwidth; gain; feed impedance; F/B and so on.

There is no 'one' cost function, and tailoring it requires some finesse.
The
tradeoff usually occurs in bandwidth compared to gain and F/B. The bigger
the
value of the cost function, the better the candidate design.

A simple expression that could be used is

CF= aG +bF +cW (1)

where G is gain in db; F is F/B in dB; and W is bandwidth in percent. the
a,b,c coefficients thus determine which is a bit more important to you,
as
well as allowing relative normalization.

73,
Chip N1IR


I was thinking that for every published or generic design maybe there could
be a kind of master list that you could consult.

ie like:
Rating
Antenna Gain F/B Bandwidth

OWA (5 ele) 3 3 5

Standard Yagi (5 ele) 4 3 2

DL6WU (10 ele) 5 6 2

DL6WU (15 ele) 6 6 2

But, you could just use the actual values I suppose. You would have in this
case three seperate tables, this one above if low to high on the gain
parameter.

I was thinking that for every published or generic design maybe there could
be a kind of master list that you could consult.

ie like:
Rating
Antenna Gain F/B Bandwidth

OWA (5 ele) 3 3 5

Standard Yagi (5 ele) 4 3 2

DL6WU (10 ele) 5 6 2

DL6WU (15 ele) 6 6 2

But, you could just use the actual values I suppose. You would have in this
case three seperate tables, this one above if low to high on the gain
parameter.



I am not aware of such a list. It would certainly be useful.

Typically, a designer has a specific need and the optimization solves for that.

73,
Chip N1IR