wrote:

Cecil Moore wrote:
Guess everyone sees the danger in trying to guess what
the results of someone else's measurement will be. Tom
should have measured something around 15.6 degrees. The
fact he didn't sends up a very large red flag.

Tom, I am going to ask you some relatively simple technical
questions. If you continue to refuse to answer those
questions, I and the other readers of r.r.a.a will draw
a logical conclusion about your unwillingness or inability
to answer questions, i.e. a non-technical answer or no
answer at all will cause you to lose credibility.

I was wrong about the radiation resistance equation. See
how readily I admitted my mistake? (When was the last time
you admitted a mistake?)

First question is a short one: Please explain why a century
old method of determining the phase shift through a coil by
measuring its self-resonant frequency is not good enough for
you. Do you really expect us to believe that the phase shift
through a well-designed coil can change by 81% from 16 MHz
to 4 MHz?

Translation of what Cecil actually is saying:

"Whenever multiple measurements by independent sources disagree with me
the measurements others made must be wrong."

I make mistakes but I seem to be on a solid technical
footing here. A number of readers agree. Maybe you can
convince me and them otherwise if you stop refusing to
answer technical questions about your measurements. Your
100 uH coil is the 8+j2500 ohm load in the following
fixed font example. That's at 3.98 MHz with a Q of 313.

Current probes are at X and Y. How is the following
circuit different from your test setup?

+---one wavelength lossless 50 ohm coax-----+
| X
source coil 8+j2500 load
| Y
+-------------------coax braid--------------+

The one wavelength of lossless coax doesn't change any values
in the steady-state situation so the current probes at X and
Y read the same value of currents as yours.

The SWR on the coax is about 16000:1. There is virtually
zero net current flowing through the load because of the
*extreme* mismatch. Virtually all of the current at the coil
is standing-wave current which is known to have unchanging
phase. Roy measured the unchanging phase of standing wave
current and reported close to zero. You measured the
unchanging phase of standing wave current and got close
to zero. It is no wonder you guys get the same value of
current phase delay since you are making exactly the same
error in your measurements spanning a number of years.
I'm surprised that you didn't measure a 0 nS delay.
--
73, Cecil http://www.qsl.net/w5dxp

Cecil, W5DXP wrote:
"Please explain why a centuries old method of determining phase shift
through a coil by measuring its self-resonant frequency is not good
enough for you?"

A coil is an RLC circuit. At resonance, L offsets C and all that is left
is R. In a resistance, the current is in-phase with the applied voltage.

But, in a physical length of a tuned circuit or in a straight conductor
in its place, in a circuit with reflections, you have energy coming from
both directions creating an interference pattern, which is repeated
every 1/2-wave (180-degrees) in the line Peaks are 1/2-wave apart,
considering the velocity factor of the line. To determine the phase
shift, count the maxima.

The wavelength of a line is the distance a wave must travel for one
complete cycle (360-degrees). If you want the phase shift for a line,
take the length of line required for one degree of phase retardation and
multiply it by the length of line you have.

Best regards, Richard Harrison, KB5WZI

Richard Harrison wrote:
The wavelength of a line is the distance a wave must travel for one
complete cycle (360-degrees). If you want the phase shift for a line,
take the length of line required for one degree of phase retardation and
multiply it by the length of line you have.

If you want to know the velocity factor of a piece of
transmission line, the easiest thing to do is find
its first self-resonant frequency. A little math
will yield the VF which allows prediction of the
phase shift through any reasonable length of
tranmission line.

If you want to know the velocity factor of a coil,
the easiest thing to do is find its first self-
resonant frequency. A little math will yield the
VF of the coil which allows prediction of the
phase shift through any reasonable length of coil.

Not disagreeing - just expanding.
--
73, Cecil http://www.qsl.net/w5dxp

Cecil Moore wrote:

If you want to know the velocity factor of a piece of
transmission line, the easiest thing to do is find
its first self-resonant frequency. A little math
will yield the VF which allows prediction of the
phase shift through any reasonable length of
tranmission line.

If you want to know the velocity factor of a coil,
the easiest thing to do is find its first self-
resonant frequency. A little math will yield the
VF of the coil which allows prediction of the
phase shift through any reasonable length of coil.

If the inductor in question does not take much advantage of mutual
induction across its length nor has much capacitance across its length
(say, a straight conductor, strung with ferrite toroids), then I can
see the similarity with a transmission line. But as the inductor
approaches a lumped inductance with significant inter winding
capacitance and mutual inductance coupling the current across a
significant part of its winding length, I see on reason to assume the
transmission line method (delay independent of frequency) strictly
applies. It might, but it would take more than you saying so to
assure me that it is a fact.

In other words, transmission line concepts like uniform inductance per
length and uniform capacitance per length get rather muddled in a real
inductor.

John Popelish wrote:

If the inductor in question does not take much advantage of mutual
induction across its length nor has much capacitance across its length
(say, a straight conductor, strung with ferrite toroids), then I can see
the similarity with a transmission line. But as the inductor approaches
a lumped inductance with significant inter winding capacitance and
mutual inductance coupling the current across a significant part of its
winding length, I see on reason to assume the transmission line method
(delay independent of frequency) strictly applies. It might, but it
would take more than you saying so to assure me that it is a fact.

In other words, transmission line concepts like uniform inductance per
length and uniform capacitance per length get rather muddled in a real
inductor.

Tom W8JI posted a good description and summary of inductor operation a
little while ago, but it looks like it could bear repeating, perhaps
with a slightly different slant.

In a transmission line, a field at one end of the line requires time to
propagate to the other end of the line. As the EM fields propagate, they
induce voltages and currents further down the line, which create their
own EM fields, and so forth. These propagating fields and the currents
and voltages they produce make the whole concept of traveling voltage
and current waves useful and meaningful.

But in a tightly wound inductor, a field created by the current in one
turn is coupled almost instantly to all the other turns (presuming that
the coil is physically very small in terms of wavelength). Consequently,
output current appears very quickly following the application of input
current. The propagation time is nowhere near the time it would take for
the current to work its way along the wire turn by turn.

Once again it's necessary to point out that I'm speaking here of an
inductor which has very good coupling between turns and minimal field
leakage or radiation, for example a toroid. If you make an air wound
inductor and slowly stretch it out until it's nothing more than a
straight wire, it'll begin by resembling the toroid -- more or less,
depending on how well coupled the turns are and how much its field
interacts with the outside world -- then slowly change its
characteristics to resemble a straight wire. There's no magic transition
point. So by choosing the inductor, you can observe behavior anywhere
along this continuum.

Roy Lewallen, W7EL

Roy Lewallen wrote:
But in a tightly wound inductor, a field created by the current in one
turn is coupled almost instantly to all the other turns ...

"All the other turns"? Here's what Jim Lux, W6RMK, had to say
about that:

"For inductance the signficant thing is that the magnetic field
of one segment pretty much links to the adjacent segments, and
less so for the rest."

Less to the 3rd, less than that to the 4th, even less than that
to the 5th. What do you think it might be by the time it gets
to the 80th turn on Tom's coil? Seems that we can assume that
the linkage between coil #1 and coil #80 is negligible.

Once again it's necessary to point out that I'm speaking here of an
inductor which has very good coupling between turns and minimal field
leakage or radiation, ...

So was W6RMK.

There's no magic transition point.

Indeed there isn't. I repeat, in case your didn't understand -
indeed there isn't. So you can discard your magic lumped-
circuit model for a system containing reflections.
--
73, Cecil http://www.qsl.net/w5dxp

John Popelish wrote:
... I see no reason to assume the transmission line method
(delay independent of frequency) strictly applies. It might, but it
would take more than you saying so to assure me that it is a fact.

Assume the environment of the coil is fixed like the variable
stinger measurement I reported earlier. Besides the frequency
term, the phase constant depends upon L, C, R, and G as does
the Z0 equation. Why would the L, C, R, and G change appreciably
over a relatively narrow frequency range as in my bugcatcher coil
measurements going from 6.7 MHz to 3.0 MHz?

And I didn't mean to imply that the delay is "independent" of
frequency, just that it is not nearly as frequency dependent
as Tom's measurements would suggest. If Tom made his measurements
from 1 MHz to 16 MHz, what do you think the curve would look like?

Freq 1 2 4 8 16 MHz
Delay ___ ___ 3 ___ 16 nS

That looks non-linear to me. How about you?
--
73, Cecil http://www.qsl.net/w5dxp

Cecil Moore wrote:
John Popelish wrote:

... I see no reason to assume the transmission line method (delay
independent of frequency) strictly applies. It might, but it would
take more than you saying so to assure me that it is a fact.


Assume the environment of the coil is fixed like the variable
stinger measurement I reported earlier. Besides the frequency
term, the phase constant depends upon L, C, R, and G as does
the Z0 equation. Why would the L, C, R, and G change appreciably
over a relatively narrow frequency range as in my bugcatcher coil
measurements going from 6.7 MHz to 3.0 MHz?

We are not talking about L, C, R, or any other inherent property
changing with frequency. We are talking about the delay of a current
wave in a single direction (anybody have a pair of directional coupler
current probes?) through a complex component that has several
different mechanisms that contribute to the total current passing
through it. It is the vector sum (superposition) of those current
components that is in question. Over a narrow frequency range, it is
conceivable to me, that the phase (delay) of that sum might shift,
dramatically, though any component of that sum might change its
magnitude only slightly (no faster than in proportion to the
frequency), and the phase of that component might change not at all.

And I didn't mean to imply that the delay is "independent" of
frequency, just that it is not nearly as frequency dependent
as Tom's measurements would suggest. If Tom made his measurements
from 1 MHz to 16 MHz, what do you think the curve would look like?

Freq 1 2 4 8 16 MHz
Delay ___ ___ 3 ___ 16 nS

That looks non-linear to me. How about you?

Definitely nonlinear, just like impedance is very nonlinear as the
frequency passes through any resonance. This is why I am suspicious
of a measurement made at resonance, being extrapolated to non resonant
conditions.

John Popelish wrote:
We are not talking about L, C, R, or any other inherent property
changing with frequency.

The velocity factor of the coil is based on those quantities
and can be calculated.

The velocity factor of a transmission line is based on those
quantities and can be calculated.

Freq 1 2 4 8 16 MHz
Delay ___ ___ 3 ___ 16 nS

That looks non-linear to me. How about you?

Definitely nonlinear, just like impedance is very nonlinear as the
frequency passes through any resonance.

Care to fill in the blanks above?

This is why I am suspicious of
a measurement made at resonance, being extrapolated to non resonant
conditions.

Self-resonance is simply where the round trip delay through
the coil puts the forward and reflected voltages and the forward
and reflected currents either at zero degrees or 180 degrees.

That's what happens at an open-ended 1/4WL stub.

That's also what happens at the feedpoint of a resonant
standing wave antenna like a 75m mobile bugcatcher antenna.
Resonant mobile antennas are "self-resonant antenna systems".
--
73, Cecil http://www.qsl.net/w5dxp