On Tue, 06 Apr 2010 00:16:56 GMT, Owen Duffy wrote:

Bob wrote in
:

...

But then, on a hunch, I checked the manual that came with my MFJ-269,
and sure enough, on page 34, it tells how to measure Velocity Factor,
utilizing the distance to fault mode. It'll take a day or so to
recharge the 269's batteries, and then I'll have at it.

As Roy has explained, you need to stop common mode current from
significantly altering your measurement.

I have had sucess with placing a balun of a string of ferrite cores over
the line. It is easy to observe the effectiveness using a VNA sweep, a
bit tricker with the MFJ269.

I do have a W2DU-style balun of ferrite beads on coax, if that is what
you mean.

I also have an MFJ gizmo, a tiny 1:1 current balun for antenna
analyzers, a coax fitting on one side, and balanced line fasteners on
the other side -- but I'm guessing then I'd be measuring the velocity
factor of the balun, in addition to the balanced line.

Bob
k5qwg


I have also found that stretching the line out straight causes the worst
common mode problems, but if you coil it, you have to keep adjacent turns
much further apart than the line's conductor separation.

All this has to be done with the line suspended in the air, well clear of
other dielectrics or conductors. (Hint: fishing line can be your friend!)

Before these analysers, we measured the resonant frequency of a line
section using a GDO. By very loosely coupling the GDO, and reading the
GDO frequency from a calibrated receiver, good results could be obtained.

Owen

Bob wrote in
:

On Tue, 06 Apr 2010 00:16:56 GMT, Owen Duffy wrote:
....
I have had sucess with placing a balun of a string of ferrite cores
over the line.

That means literally threading some suitable ferrite toroidal cores over
the transmission line you are measuring.

If you add a separate balun between the analyser and the cable under test,
you introduce an unknown component that will probably disturb your
readings.

Owen

On Tue, 06 Apr 2010 01:26:16 GMT, Owen Duffy wrote:

Bob wrote in
:

On Tue, 06 Apr 2010 00:16:56 GMT, Owen Duffy wrote:
...
I have had sucess with placing a balun of a string of ferrite cores
over the line.

That means literally threading some suitable ferrite toroidal cores over
the transmission line you are measuring.

If you add a separate balun between the analyser and the cable under test,
you introduce an unknown component that will probably disturb your
readings.

Owen

Another question -- I'm thinking of cutting a 10-foot section of
balanced line to test. Should I count the bared pigtails of the line,
which I will attach to the analyzer's coax output, as part of the 10
foot length? Or just count that part of the line where all insulation
is in place?

Bob
k5qwg

On Apr 5, 3:33*pm, Owen Duffy wrote:
Typical T match ATU's are lossier on capacitive loads than on inductive
loads.

How about typical CLC Pi-Net ATUs?
--
73, Cecil, w5dxp.com

Bob wrote:
On Tue, 06 Apr 2010 01:26:16 GMT, Owen Duffy wrote:

Bob wrote in
:

On Tue, 06 Apr 2010 00:16:56 GMT, Owen Duffy wrote:
...
I have had sucess with placing a balun of a string of ferrite cores
over the line.
That means literally threading some suitable ferrite toroidal cores over
the transmission line you are measuring.

If you add a separate balun between the analyser and the cable under test,
you introduce an unknown component that will probably disturb your
readings.

Owen

Another question -- I'm thinking of cutting a 10-foot section of
balanced line to test. Should I count the bared pigtails of the line,
which I will attach to the analyzer's coax output, as part of the 10
foot length? Or just count that part of the line where all insulation
is in place?


Aha.. you start to see the problems in precision RF measurement... Where
is the "reference plane"..and how do you calibrate out the "fixture".

One way to do it is to do two sets of measurements. Do one with your 10
foot length. Then, cut 5 feet off and do it again. Then, the
"difference" between the measurements is the result for the now missing
5 feet.

How much precision are you looking for, anyway. To a first order, think
about how long that fixture is. If it's an inch or so, that's less than
1% of the overall length of the line.

Bob wrote:

Another question -- I'm thinking of cutting a 10-foot section of
balanced line to test. Should I count the bared pigtails of the line,
which I will attach to the analyzer's coax output, as part of the 10
foot length? Or just count that part of the line where all insulation
is in place?

Bob
k5qwg

I think 10 feet is going to be too short to make a good measurement,
because the lengths of such things as the pigtails and the MFJ are a
substantial fraction of the overall length. I recommend using the whole
length of line you have. You might have to be a bit creative in keeping
it away from other conductors, but that'll give you the best results.

When you do make the measurement, maintain the integrity of the line to
as close to the impedance meter as you can. Then measure the line to the
impedance meter connector.

Roy Lewallen, W7EL

Bob wrote in
:

....
Another question -- I'm thinking of cutting a 10-foot section of
balanced line to test. Should I count the bared pigtails of the line,
which I will attach to the analyzer's coax output, as part of the 10
foot length? Or just count that part of the line where all insulation
is in place?

What you have is two transmission line sections in cascade, one with bare
conductors, and one with the conductors immersed in insulation.

If you want to measure the effects only of the latter, you need to find
some way of minimising the contribution of the former.

The calibration of the MFJ269 is not that flash that you will pick a mm
or two. When I have used them for the test you are performing, I zip tie
the conductor to the external threads of the connector so that there is
as close to zero length of 'different' transmission line as possible. You
could also use a small stainless hose clamp, but in my experience, the
zip tie has been reliable.

You can zip tie a piece of PE irrigation pipe to the VFO knob so that you
hand doesn't need to be within half a meter of the instrument, use a
wooden table to support the instrument, use the balun I suggested, and
arrange the line to minimise radiation from residual common mode current.

I would try to measure a length of 10m or so. It is a compromise between
making end effects (tails, effect of the windows) insignificant, an
effective balun, and physically supporting the line for least radiation
and other external influences.

Some of my focus was on trying to get a valid measure of R as well as X,
R due to line losses alone.

Owen

On Tue, 06 Apr 2010 10:11:20 -0700, Roy Lewallen
wrote:

Bob wrote:

Another question -- I'm thinking of cutting a 10-foot section of
balanced line to test. Should I count the bared pigtails of the line,
which I will attach to the analyzer's coax output, as part of the 10
foot length? Or just count that part of the line where all insulation
is in place?

Bob
k5qwg

I think 10 feet is going to be too short to make a good measurement,
because the lengths of such things as the pigtails and the MFJ are a
substantial fraction of the overall length. I recommend using the whole
length of line you have. You might have to be a bit creative in keeping
it away from other conductors, but that'll give you the best results.

When you do make the measurement, maintain the integrity of the line to
as close to the impedance meter as you can. Then measure the line to the
impedance meter connector.

Roy Lewallen, W7EL

I have 53-foot- and 122-foot-long lengths of the line. I might stretch
the 53-footer from the roof out toward the back fench, and measure
that.

Bob
k5qwg

Roy Lewallen Inscribed thus:

Baron wrote:

Please could you elaborate on how and why a common mode current has a
different VF on a balanced line.

Sure.

First, a balanced line, whether it's twinlead or coax, doesn't have
any common mode current, by definition -- the lack of common mode is
what makes it balanced. We're talking about a physically symmetrical
line.

Whenever you have a two conductor line, you effectively have two
transmission lines, differential mode and common mode. Although you
actually have only one current on each conductor, by taking advantage
of the principle of superposition you can mathematically separate the
two currents into two *sets* or components of currents, analyze their
effects separately to gain a better understanding, and simply add the
results if you want to know the overall solution. The sum of the
common mode and differential currents are the actual conductor
currents, and the sum of the common mode and differential responses is
the actual response.

The differential or transmission line mode waves (voltage and current)
are the components which are equal and opposite on the two conductors,
so the field is strongest between the two conductors, fringing outward
in the case of ladder line. The presence of the dielectric material in
a major portion of the field slows down the waves, lowering the
velocity factor. In the case of coax, the field is entirely within the
dielectric so we can easily calculate the velocity factor if we know
the dielectric constant of the material. In the case of ladder line,
we don't know what fraction of the field is in the air and what's in
the dielectric without a very advanced computer program, so we have to
measure the velocity factor. The fraction and therefore velocity
factor changes, by the way, with frequency, a phenomenon known as
dispersion.

The common or antenna mode waves are the components that are equal and
in the same direction or polarity on the two conductors. The field is
the same as it would be if the two conductors were connected together
to make a single conductor. One conductor of the common mode
transmission line is the two conductors of the ladder line, and the
other is the Earth and/or surrounding conductors. These two common
mode transmission line conductors are usually much farther apart than
the ladder line conductors, so the common mode characteristic
impedance is higher than the differential mode impedance. The velocity
factor is usually higher, too, because the field is between the two
common mode conductors -- the ladder line and the Earth --, and almost
none of it is in the line dielectric. So its velocity factor is nearly
1. In my TDR demonstration, the common mode open end reflection
occurred before the larger differential mode reflection because of the
higher velocity factor, so it looked like a differential mode
reflection from a point short of the end. (And I helped reinforce this
mistake in order to get the audience's attention.)

Any two conductor line supports both modes and behave the same, but
coax is a little easier to understand because the differential and
common mode currents are actually physically separate -- so no
mathematical hocus-pocus is necessary. The differential currents and
waves are entirely inside the cable, and the common mode currents and
waves are outside. The velocity factor inside (differential mode) is
determined by the dielectric material, and the velocity factor of the
outside (common mode) is nearly 1.

Roy Lewallen, W7EL

Thankyou, Jim & Roy.
Your explanations were most enlightening. I just couldn't get my head
around the "how & why" the VF should be different. I have also
realised why I have sometimes seen more than one TDR reflection from a
perfectly good transmission line.

73's
--
Best Regards:
Baron.

Baron wrote:

Thankyou, Jim & Roy.
Your explanations were most enlightening. I just couldn't get my head
around the "how & why" the VF should be different. I have also
realised why I have sometimes seen more than one TDR reflection from a
perfectly good transmission line.

You can easily excite a common mode wave on coax with a TDR -- or
transmitter -- simply by connecting to it with pigtails. This provides a
path between the inside and outside of the shield, unlike a proper coax
connector which preserves the integrity of the shield.

Roy Lewallen, W7EL