23cm 47-element G3JVL loop-Yagi gain and beamwidth figures

[Andrew Marshall]

I have a 23cm G3JVL 47-element loop-Yagi at 24ft AGL. I have not so far
been able to find any gain and beamwidth figures for it by searching
online. Has anyone got these figures to hand, please?

TIA.

--
Regards,
Andrew, G8BUR, M0MAA.

[Brian Howie]

uk!brian
References: <rrsqm7$d9s$1...@dont-email.me>
Lines: 22
MIME-Version: 1.0
Content-Type: text/plain;charset=us-ascii;format=flowed
User-Agent: Turnpike/6.07-S (<GnW4jgrj$LIyz1x6IZKPnfBglz>)
To: uk-radio-amat...@moderators.isc.org
X-Antivirus: Avast (VPS 201222-0, 22/12/2020), Outbound message
X-Antivirus-Status: Clean

In m
essage <rrsqm7$d9s$1...@dont-email.me>, Andrew Marshall
<ne...@y-ddraenog-goch.org.uk> writes

>I have a 23cm G3JVL 47-element loop-Yagi at 24ft AGL. I have not so far
>been able to find any gain and beamwidth figures for it by searching
>online. Has any
one got these figures to hand, please?
>
>TIA.
>

I've a model of a 38 ele loop yagi. Extending it to 47 elements and
running the model gives a gain of 21.56dBi and a 3dB beam width of
between 14 and 16 degrees . The F/B is 24dB.

There's a com
mercial 49 ele here claiming 21.3dbi and 15 deg B/W

http://www.vpa-systems.pl/1296mhz-loop-yagi-49el-213dbi-400cm-p-167.html

Cheers

Brian GM4DIJ
--
Brian Howie

[Andrew Marshall]

On 22/12/2020 14:39, Brian Howie wrote:
>> In message <rrsqm7$d9s$1...@dont-email.me>, Andrew Marshall

>> <ne...@y-ddraenog-goch.org.uk> writes
>> I have a 23cm G3JVL 47-element loop-Yagi at 24ft AGL. I have not so far
>> been able to find any gain and beamwidth figures for it by searching
>> online. Has any
>> one got these figures to hand, please?

>> TIA.

> I've a model of a 38 ele loop yagi. Extending it to 47 elements and
> running the model gives a gain of 21.56dBi and a 3dB beam width of
> between 14 and 16 degrees . The F/B is 24dB.

Excellent - thank you very much. That ought to give me an adequately
accurate basis for assessing ICNIRP compliance. With 16 degrees 3dB
beamwidth, then to a person on the ground, there should be very little
incident field strength - I'll only be running the basic 10W RF out of
the IC-9700, less the loss of about 7.5m of Pope H100.

> There's a com
> mercial 49 ele here claiming 21.3dbi and 15 deg B/W

> http://www.vpa-systems.pl/1296mhz-loop-yagi-49el-213dbi-400cm-p-167.html

(looks) That seems to suggest that 21.5dBi would be a good figure to
work with for now. I'll point it at the 'official' calculator and see
what 'safe distance' is returned. I only wish that it were not so
conservative as it is reported to be.

All my other antennas seem to be compliant, except for the 'random'
~130m-long horizontal loop at around 4m AGL running around the garden,
which wouldn't appear to radiate much at a downwards angle of 20 to 30
degrees or so to a person on the ground - according to a model of an 80m
full-wave loop in the fairly-recent RSGB 'antennas' book, it pretty much
all goes upwards, which is what I wanted, for 40/80m NVIS working, when
I put it up many years ago.

Many thanks for the modelling and the link.

--
73,
Andrew, G8BUR, M0MAA.

[Andrew Marshall]

On 24/12/2020 09:31, Jeff wrote:
>  > (looks) That seems to suggest that 21.5dBi would be a good figure to
>> work with for now. I'll point it at the 'official' calculator and see
>> what 'safe distance' is returned. I only wish that it were not so
>> conservative as it is reported to be.
>>
>> All my other antennas seem to be compliant, except for the 'random'
>> ~130m-long horizontal loop at around 4m AGL running around the garden,
>> which wouldn't appear to radiate much at a downwards angle of 20 to 30
>> degrees or so to a person on the ground - according to a model of an
>> 80m full-wave loop in the fairly-recent RSGB 'antennas' book, it
>> pretty much all goes upwards, which is what I wanted, for 40/80m NVIS
>> working, when I put it up many years ago.
>>
>> Many thanks for the modelling and the link.

> For the loop any members of the 'public' are almost certainly going to
> be in the near field of the antenna at HF, so it will be the near field
> E and H fields that you have to consider for ICNIRP compliance; which at
> close distances may well be very different from the resultant radiation
> pattern.
>
> The best way to do that would be to simulate it using a program such as
> 4NEC2 which will give you those figures (graphically if required).

Thank you for that info.

I hadn't seen any mention of near or far fields on the Ofcom calculator
spreadsheet, so I thought that to use it as a compliance demonstrator
would suffice. If that isn't the case, I will have to suspend use of
that antenna until I can find a way of demonstrating compliance in the
near field (unless there is a version of 4NEC2 or other program that
runs under Linux Mint, I will be in some difficulty).

I can only imagine that thousands of HF operators will be in the same
boat as me in this respect.

--
73,
Andrew.

[Brian Howie]

In message <rs4a5p$rhg$1...@dont-email.me>, Andrew Marshall
<ne...@y-ddraenog-goch.org.uk> writes

4NEC2 will run under Wine on Linux Mint. The Offcom ,model is far-field
only. It is therefore over pessimistic for short distances for HF
anyway.

It also includes ground-gain , which is only relevant in the far-field.
I tried comparing it with a hand calculation and the VK3UM model and
discovered this fact. The VK3UM model has a ground gain option.

[Andrew Marshall]

I have a spare laptop running Linux, though I haven't used it for a year
or so. If it still works, I may try downloading and installing Wine and
then 4NEC2 (but see below). I don't want to risk rendering one of my
main Linux boxes unusable because of some installation disaster,
especially as their availability is critically important at the moment
for online food shopping and other important tasks.

> The Offcom ,model is far-field
> only. It is therefore over pessimistic  for short distances for HF anyway.

Thank you for that info. In that case, it looks as if I may be OK using
the Ofcom calculator for HF, if I can reasonably use the model of the
80m full-wave loop in 'HF Antennas for Everyone' (RSGB; p221, fig. 5),
where radiation at 0 degrees appears to be minimal (according to the
small graph, apparently somewhere between -10 and -30dB below the peak
quoted gain of 8.3dBi). Allowing a 50% duty cycle should help, too.

> It also includes ground-gain , which is only relevant in the far-field.
> I tried comparing it with a hand calculation and the VK3UM model and
> discovered this fact.  The VK3UM model has a ground gain option.

It'll certainly be an interesting modelling to carry out when
circumstances permit. If for any reason I do feel it necessary to
suspend operations using the loop, I will have to do so, but as most of
my operation is above 30MHz, where it seems that all my power/antenna
options will be compliant, it will be no great problem.

I can see a fair few second-hand linear amplifiers coming onto the
market at keen prices before long, though!

Thank you for the information.

[Paul]

Andrew Marshall wrote:
> On 25/12/2020 14:07, Brian Howie wrote:
>
>> 4NEC2 will run under Wine on Linux Mint.
>
> I have a spare laptop running Linux, though I haven't used it for a year
> or so. If it still works, I may try downloading and installing Wine and
> then 4NEC2 (but see below). I don't want to risk rendering one of my
> main Linux boxes unusable because of some installation disaster,
> especially as their availability is critically important at the moment
> for online food shopping and other important tasks.

Take a spare hard drive and clone over the working disk drive.

With a Linux fleet as important as the one you describe, you
must already be familiar with the drill.

That's how I ended up with two dozen hard drives in this room.
Any time I lack confidence in the outcome of a experiment, it
takes an extra ten minutes to prepare for it. For example, I
have a terabyte of downloads, and that is backed up on a much
less frequent basis. The tiny OS partitions can be backed up or
cloned in ten minutes or so.

On Linux, there is CloneZilla. Boot a USB stick, then do the
drive-to-drive transfer. On Windows I use Macrium Reflect Free,
which handles both NTFS and EXT4 partitions, and is "good enough"
for the Linux installs I do. The current drive in the Test Machine
has 23 partitions at the moment (and Macrium can back up the
whole thing, using just as smart a transfer method as Clonezilla).
I can't really add any more partitions, because the screen is
no longer wide enough. GPT allows 128 partitions, but then you'd
not have a big enough screen to work on them.

Paul

[Brian Howie]

It's been a while since I searched for Linux NEC2 programs.

I found xnec2 which is a native Linux version of nec2.

https://www.linuxjournal.com/content/antennas-linux

The debian install command runs fine under Linux Mint. I only got as far
as running the examples, but it seems to work.

[Andrew Marshall]

On 26/12/2020 14:13, Brian Howie wrote:
> It's been a while since I searched for Linux NEC2 programs.

> I found xnec2 which is a native Linux version of nec2.

> https://www.linuxjournal.com/content/antennas-linux

> The debian install command runs fine under Linux Mint. I only got as far
> as running the examples, but it seems to work.

Many thanks for the link. I'll dig out the old laptop, and if it works
I'll download xnec2 and try it out. I suspect that I might wait a bit
first, until I can track down more info on the radiation patterns of
NVIS antennas, and perhaps consider whether my installation might
benefit from a change anyway; NVIS, though interesting, is but a small
part of my operating, so I'll take things at a measured pace.

Thank you very much for your help and suggestions.

[Andrew Marshall]

On 26/12/2020 18:45, Paul wrote:
> Andrew Marshall wrote:
>> On 25/12/2020 14:07, Brian Howie wrote:
>>
>>> 4NEC2 will  run under Wine on Linux Mint.
>>
>> I have a spare laptop running Linux, though I haven't used it for a
>> year or so. If it still works, I may try downloading and installing
>> Wine and then 4NEC2 (but see below). I don't want to risk rendering
>> one of my main Linux boxes unusable because of some installation
>> disaster, especially as their availability is critically important at
>> the moment for online food shopping and other important tasks.
>
> Take a spare hard drive and clone over the working disk drive.

That I can easily do. I have a couple of portable HDDs which ought to be
big enough.

> With a Linux fleet as important as the one you describe, you
> must already be familiar with the drill.

Shouldn't be a problem.

> That's how I ended up with two dozen hard drives in this room.
> Any time I lack confidence in the outcome of a experiment, it
> takes an extra ten minutes to prepare for it. For example, I
> have a terabyte of downloads, and that is backed up on a much
> less frequent basis. The tiny OS partitions can be backed up or
> cloned in ten minutes or so.

I have nothing like that amount so my backups don't take very long.

> On Linux, there is CloneZilla. Boot a USB stick, then do the
> drive-to-drive transfer. On Windows I use Macrium Reflect Free,
> which handles both NTFS and EXT4 partitions, and is "good enough"
> for the Linux installs I do. The current drive in the Test Machine
> has 23 partitions at the moment (and Macrium can back up the
> whole thing, using just as smart a transfer method as Clonezilla).
> I can't really add any more partitions, because the screen is
> no longer wide enough. GPT allows 128 partitions, but then you'd
> not have a big enough screen to work on them.

Thank you for the above (of which I have taken a copy).

[David Woolley]

On 24/12/2020 09:31, Jeff wrote:
> For the loop any members of the 'public' are almost certainly going to
> be in the near field of the antenna at HF, so it will be the near field
> E and H fields that you have to consider for ICNIRP compliance; which at
> close distances may well be very different from the resultant radiation
> pattern.
>

It is also going to be dependant on frequency and ground conductivity.
Unless you have a really good ground mesh, extending well beyond the
footprint, you are going to need to model for different ground
conductivities, to allow for unknowns and weather effects, as well as at
different frequencies.

Also, far field polar diagrams are calculated for very large distances,
where the angle subtended by the antenna will be negligible compared
with the beam width, and you can assume all the radiation comes from a
central point. Near field levels may even be higher near the actual
wire, than in the centre of the antenna.

[Andrew Marshall]

On 27/12/2020 11:56, Jeff wrote:
>
>>> The Offcom ,model is far-field only. It is therefore over pessimistic
>>> for short distances for HF anyway.
>>
>> Thank you for that info. In that case, it looks as if I may be OK
>> using the Ofcom calculator for HF, if I can reasonably use the model
>> of the 80m full-wave loop in 'HF Antennas for Everyone' (RSGB; p221,
>> fig. 5), where radiation at 0 degrees appears to be minimal (according
>> to the small graph, apparently somewhere between -10 and -30dB below
>> the peak quoted gain of 8.3dBi). Allowing a 50% duty cycle should
>> help, too.
>>
>

> I think that is a false assumption; in the near field the patterns can
> be remarkably different from the far field pattern.
>
> When close to a wire the near E & H fields will look nothing like
> resultant fields at a distance, at 0 degrees there will be strong fields.

I'm thinking more and more that I (and many others) will have to wait
for more precise information on what analysis of radiation patterns will
be required to demonstrate compliance when using HF with wire antennas,
before using them in future.