wrote:
Why worry about VHF though?

Because MFJ advertises 2m operation with the
recommended coil of coax?
--
73, Cecil http://www.w5dxp.com

Cecil Moore wrote:
Ian White GM3SEK wrote:

[Snip]

Only one part of that posting grabbed my attention:


beaded 6 36 6.0

Beaded isn't a coil so doesn't count.


The word "beaded" did not exist in my original posting... but there it
is in Cecil's reply, complete with the double that attributes it to
me, so that Cecil can knock the straw-man down.


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

Cecil Moore wrote:
Ian White GM3SEK wrote:
Across the whole 1-30MHz band, the bunched choke behaves as an almost
perfect L-C circuit, free from any unwanted resonances.

One additional point. If the above were true, as the
frequency is increased, the phase angle of the coiled
choke impedance would drop from ~90 degrees to zero at
the self-resonant frequency, and then rise back to ~-90
degrees and stay there. But that's not what happens.

In every single case, the phase angle rises toward -90
degrees *and then decreases* as the 1/2WL self-resonance
point is approached. That is a clear indication of transmission
line effects. A lumped circuit simply doesn't act that way.

The spreadsheet at
http://www.ifwtech.co.uk/g3sek/misc/chokes.xls
now includes graphs of the phase information for all the chokes.

[ Thank you, Owen - the phase axis label is now fixed .
For anyone interested who doesn't have Excel, Microsoft's free Excel
file viewer is at http://tinyurl.com/cup85 ]

Coming all the way back to the original question, the data confirms that
even though it may look like something held together with duct tape, a
coiled coax choke can be an excellent single-band solution. At its
parallel self-resonant frequency, it will have a much higher common-mode
impedance than a generic string of ferrite beads. It can probably
outperform or at least equal a ferrite choke over two or possibly three
adjacent HF bands; but being a resonant device, it cannot deliver
extreme broadband performance.

In the chokes we're looking at, the low-impedance series resonances of
which Cecil complains do not occur below 30MHz. Those resonances exist,
but not on the HF frequencies where the chokes would actually be used.

Within the practical working frequency range of all of these coiled-coax
chokes, the performance can be accurately described as that of a simple
parallel tuned LC circuit, which displays no transmission-line behaviour
whatever. Cecil complains that
One might say it is misbehaving and is a very poor design.

That sounds to me like the complaint of someone who has a pet theory to
hammer, and is disappointed when he can't find a nail.

I think that's it, really. The graphs themselves say the rest.


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

On Sun, 22 Oct 2006 23:56:04 +0100, Ian White GM3SEK
wrote:

[ Thank you, Owen - the phase axis label is now fixed .
For anyone interested who doesn't have Excel, Microsoft's free Excel
file viewer is at http://tinyurl.com/cup85 ]

Hi Owen,

It is far easier to simply use OpenOffice which is an executable that
will translate to/from Windows Office (any spread sheet, document,
presentation, drawing...). The Open Document format is the mandated
standard of the European Community, if I recall correctly.

73's
Richard Clark, KB7QHC

On Sun, 22 Oct 2006 16:24:33 -0700, Richard Clark
wrote:

Hi Owen,

Sorry Ian, wrong side of the continental divide here.

Ian White GM3SEK wrote:
beaded 6 36 6.0
Beaded isn't a coil so doesn't count.

The word "beaded" did not exist in my original posting... but there it
is in Cecil's reply, complete with the double that attributes it to
me, so that Cecil can knock the straw-man down.

Ian, I think I corrected an obvious typo of yours. It
appeared that you had gotten off by one row. If that
was wrong, I apologize. Here's what you posted:

Choke Fmax Fmin Ratio
6t 1 layer 24 none -
12t 1 layer 15 31 2.1
4t 1 layer 21 34 1.6
8t 1 layer 12 19 1.6
12 32 2.7
8t bunched 6 36 6.0

There were six chokes. The fifth was bunched and the
sixth was beaded. The bunched ratio is 2.7. It seemed
obvious that you had made a typo in your chart which
I corrected.
--
73, Cecil http://www.w5dxp.com

Ian White GM3SEK wrote:
In the chokes we're looking at, the low-impedance series resonances of
which Cecil complains do not occur below 30MHz. Those resonances exist,
but not on the HF frequencies where the chokes would actually be used.

But, Ian, what about 75m and 40m?

Those resonances above 26 MHz are transmission line effects.
The largest coil is only 12 turns, not nearly enough for 75m
operation. The largest 12 turn coil has a choking impedance
of only 100 ohms on 75m. Scale the choke to 75m and 40-50 turns
are probably required. That would certainly lower the 1/2WL
resonant frequencies into the HF ham bands.

The 12 turn coil's peak impedance is around 15 MHz and its
1/2WL low impedance response is around 31 MHz. If we scale
the peak impedance value to 4 MHz by increasing the turns,
the 1/2WL low impedance response would be around 8 MHz. Such
a choke would be useless above 10 MHz.

Bottom Line: If enough turns are used to achieve a high choking
impedance on 75m, the 1/2WL resonant transmission line effects
will occur in the middle of HF making it useless on the higher
HF frequencies.

Within the practical working frequency range of all of these coiled-coax
chokes, the performance can be accurately described as that of a simple
parallel tuned LC circuit, which displays no transmission-line behaviour
whatever.

This is obviously a false statement proved by your own graphs.
If there were no transmission line effects, the phase graphs
would all converge on -90 degrees. All of the phase bumps above
26 MHz are transmission line effects. If the chokes had their
maximum impedance at 4 MHz, those 1/2WL transmission line effects
would probably be in the range of 10 MHz rendering them useless
on 30m-10m.

Cecil complains that
One might say it is misbehaving and is a very poor design.

That sounds to me like the complaint of someone who has a pet theory to
hammer, and is disappointed when he can't find a nail.

Actually, I was just quoting this web page:

http://www.k1ttt.net/technote/airbalun.html

"- Don't bunch the turns together. Wind them as a single
layer on a form. Bunching the turns kills the choking effect
at higher frequencies." I agree with WA2SRQ that the bunched
coil choke is a poor design. Do you think that WA2SRQ has a
pet theory?

I think that's it, really. The graphs themselves say the rest.

They certainly do. Everything above 26 MHz is obviously transmission
line effects. If the chokes were designed for the lower part
of the HF spectrum, the 1/2WL low impedance points, caused by
the transmission line effects, would be in the middle of the HF
spectrum.

Scale the high impedance points for 75m and the low impedance
bumps due to transmission line effects will occur in the
middle of HF.
--
73, Cecil http://www.w5dxp.com

On Sun, 22 Oct 2006 16:24:33 -0700, Richard Clark
wrote:

Hi Owen,

It is far easier to simply use OpenOffice which is an executable that
will translate to/from Windows Office (any spread sheet, document,
presentation, drawing...). The Open Document format is the mandated
standard of the European Community, if I recall correctly.

Be careful. If any VB type macros are included in Excel Open Office
will barf. Been there done that.

Danny, K6MHE

Cecil Moore wrote:
Ian White GM3SEK wrote:
beaded 6 36 6.0
Beaded isn't a coil so doesn't count.

The word "beaded" did not exist in my original posting... but there
it is in Cecil's reply, complete with the double that attributes it
me, so that Cecil can knock the straw-man down.

Ian, I think I corrected an obvious typo of yours. It
appeared that you had gotten off by one row. If that
was wrong, I apologize. Here's what you posted:

Choke Fmax Fmin Ratio
6t 1 layer 24 none -
12t 1 layer 15 31 2.1
4t 1 layer 21 34 1.6
8t 1 layer 12 19 1.6
12 32 2.7
8t bunched 6 36 6.0

There were six chokes. The fifth was bunched and the
sixth was beaded. The bunched ratio is 2.7. It seemed
obvious that you had made a typo in your chart which
I corrected.

Apology accepted. What you had missed was the statement immediately
above the table, that:
(the 8t 1 layer choke has two very small minima).
Therefore that choke has two lines of data.

The bunched choke is the one that has a Fmax/Fmin frequency ratio of 6.
A certain amount of mental gymnastics allowed you to claim that 1.5, 2.1
and 2.7 are all approximately equal to 2... so how about 6?


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

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Danny Richardson wrote:
On Sun, 22 Oct 2006 16:24:33 -0700, Richard Clark
wrote:

Hi Owen,

It is far easier to simply use OpenOffice which is an executable that
will translate to/from Windows Office (any spread sheet, document,
presentation, drawing...). The Open Document format is the mandated
standard of the European Community, if I recall correctly.

Be careful. If any VB type macros are included in Excel Open Office
will barf. Been there done that.

Danny, K6MHE



Opens fine in OpenOffice 2 on SUSE 10 but I did have to reposition
the graphs which were over the table.

Charlie.

- --
www.wymsey.co.uk
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org

iD8DBQFFPIKufhKqaKFAJ/YRAgnOAJ9ZjDMfv6s+TwcswZd0YbJQJK5e1QCcCskT
Dbm7fF6MOISjjCsbzjBLihA=
=zEFh
-----END PGP SIGNATURE-----