K7ITM wrote:
cascaded. Each was 1uH series, followed by 100pF shunt to ground. I
put a 100 ohm load on one end and fed the other end with a 2.5MHz sine
wave with 100 ohms source resistance. Sqrt(LC) is 10 nanoseconds per
section, so I expect 100 nanoseconds total delay, or 90 degrees at
2.5MHz. That's what I saw. Then I added unity coupling among all the
coils, and to keep the same net inductance, I decreased each inductor
to 100nH. The result was STILL very close to a 90 degree phase shift,
with a small loss in amplitude. In each case, the current in each
successive inductor shifts phase by about 1/10 the total. Although the
simulation is less than a perfect match to a completely distributed
system with perfect flux linkage (and just how you do that I'm not
quite sure anyway...), but it's close enough to convince me that
perfect flux linkage would not prevent behaviour like a transmission
line, given the requisite distributed capacitance.

Thanks Tom,

That's very interesting. My thought is the difference in
phase-of-current at each of the inductor would be affected by mutual
coupling, with perfect coupling preventing phase differences in
current, but maybe that is shortsighted.

I'll have to think about that a while and how it might affect what I am
saying.

73 Tom

The phase shift in degrees, along a coil or any other sort of
transmission line, is fixed rigidly by its physical dimensions and
test frequency.

Phase shift is entirely independent of the way it is used, the circuit
it is in and the circuit currents which flow.
----
Reg. G4FGQ

Of course, that's blatantly false, taken literally. A 2" diameter 10"
long solenoid coil coaxially inside a 2.5" ID grounded conductive tube
will not have the same phase shift as the identical coil inside a 5" ID
grounded conductive tube, and neither will behave the same as the same
coil included as a loading coil in Cecil's mobile antenna. It won't
even have the same inductance in each case.

Before you say, "Give us a break, Tom. Of course it won't and clearly
that's not what was meant," just consider how literally both the
posters and the lurkers here take things.

AND in fact, as shown in the simulation I just reported on, the
coupling between that coil and the magnetic fields of other nearby
components does affect the performance of that coil. In general, when
the fields, electric and magnetic, around any component interact with
their environment, a change in that environment will change the
behaviour of the component. Thankfully, we have a lot of components
where that effect is minimal at the frequencies of interest, but we do
need to take note of cases where the effect is important. I DAILY work
with tiny components that DO behave differently, depending on their
environment. At several GHz, seemingly small couplings can be very
important.

Cheers,
Tom

wrote:
I'll have to think about that a while and how it might affect what I am
saying.

I'm back from Tulsa and had time to think while I was gone.
Given that none of the measurements reported so far actually
measured the phase shift through a coil, I have devised an
EZNEC example that should be very easy to duplicate for
real world measurements.

Previously I had offered a 5.89 MHz base-loaded antenna as
an example. The top of my antenna was the typical open-
circuit with its 100% reflection at the tip. W7EL took my
file and connected the top of the coil back to ground. He
changed the 'tip' of the antenna to a short-circuit to
ground so the 100% reflection at the 'tip' remained. All
that shorting the top of the coil to ground accomplished
was different phase shifts in the reflected wave. Any
information contained in the standing wave current phase
continued to be zero because of unchanging phase.

So here's the EZNEC example and an experiment that any
properly equipped person can duplicate. That includes
you and W7EL.

I took W7EL's EZNEC file and changed wire #203 from 0.25'
to 31.25'. At the 'tip' of the antenna, I installed a
439.2 ohm load that turns the antenna into a 90 degree
long *traveling-wave* antenna. Note that the current
magnitude at the top of the coil is identical to the
current magnitude through the load resistor. The load
resistor's value is very close to the calculated Z0
of the 31' #16 wire two feet above ground, using the
formula for a single wire transmission line above
ground.

The graphic is at http://www.qsl.net/w5dxp/test316y.GIF

The EZNEC file can be downloaded from:

http://www qsl.net/w5dxp/test316y.EZ

I will add the supporting text to my web page later today.

Please explain the 15.68 degree phase shift through the
coil. Don't you find it strange that all the wire in the
system occupies 90.01 - 15.68 = 74.33 degrees?
--
73, Cecil http://www.qsl.net/w5dxp

Cecil Moore wrote:
wrote:

I'll have to think about that a while and how it might affect what I am
saying.
(snip)
So here's the EZNEC example and an experiment that any
properly equipped person can duplicate. That includes
you and W7EL.

I took W7EL's EZNEC file and changed wire #203 from 0.25'
to 31.25'. At the 'tip' of the antenna, I installed a
439.2 ohm load that turns the antenna into a 90 degree
long *traveling-wave* antenna. Note that the current
magnitude at the top of the coil is identical to the
current magnitude through the load resistor. The load
resistor's value is very close to the calculated Z0
of the 31' #16 wire two feet above ground, using the
formula for a single wire transmission line above
ground.

The graphic is at http://www.qsl.net/w5dxp/test316y.GIF

The EZNEC file can be downloaded from:

http://www qsl.net/w5dxp/test316y.EZ
(snip)

Excellent!

Can you use this example, with varying frequency to explore your
assertion that the time delay (frequency times phase shift) of the
coil varies little over a significant range of frequencies up to self
resonance, and that that delay is about 1/4 cycle of the self resonant
frequency?

A graph of delay versus frequency would be useful. It should show
over what frequency range the coil acts mostly like a transmission
line and where it acts mostly like something else (i.e. inductor,
parallel resonant tank).

John Popelish wrote:
Can you use this example, with varying frequency to explore your
assertion that the time delay (frequency times phase shift) of the coil
varies little over a significant range of frequencies up to self
resonance, and that that delay is about 1/4 cycle of the self resonant
frequency?

I will do that when my energy level returns after getting
home at 2 am this morning. Note that anyone can download
the EZNEC file from http://www.qsl.net/w5dxp/test316y.EZ

A graph of delay versus frequency would be useful. It should show over
what frequency range the coil acts mostly like a transmission line and
where it acts mostly like something else (i.e. inductor, parallel
resonant tank).

This coil, operated below its self-resonant frequency, has
phase shift of 15.68 degrees or ~0.044 wavelength (delay of
7.4 nS). Dr. Corum says anything over 15 degrees requires
the distributed network model. 15 degrees will transform
50 ohms to 54+j120 ohms, causing SWR to be erroneously
reported as 7:1 instead of 1:1. That sounds like too
large an error to me.

Since the lumped-circuit model assumes a delay of zero, i.e.
faster than light, seems the use of the lumped-circuit model
results in 100% error, or infinite error if one calculates
it the other way. :-)

BTW, one of the principles on the other side of the argument
sent me a file with a worm in it. I guess he wanted to
extend the silence caused by my trip by bringing down my
computer.
--
73, Cecil http://www.qsl.net/w5dxp

Cecil Moore wrote:

This coil, operated below its self-resonant frequency, has
phase shift of 15.68 degrees or ~0.044 wavelength (delay of
7.4 nS). Dr. Corum says anything over 15 degrees requires
the distributed network model. 15 degrees will transform
50 ohms to 54+j120 ohms, causing SWR to be erroneously
reported as 7:1 instead of 1:1. That sounds like too
large an error to me.

Since the lumped-circuit model assumes a delay of zero, i.e.
faster than light, seems the use of the lumped-circuit model
results in 100% error, or infinite error if one calculates
it the other way. :-)

Not if the lumped inductor model includes lumps of capacitance that
represent the strays to ground. Lumped LC networks exhibit phase
shift, also.

BTW, one of the principles on the other side of the argument
sent me a file with a worm in it. I guess he wanted to
extend the silence caused by my trip by bringing down my
computer.

Never blame malice when ignorance will suffice. Even if you are
wrong, you will sleep better.

John Popelish wrote:
Not if the lumped inductor model includes lumps of capacitance that
represent the strays to ground. Lumped LC networks exhibit phase shift,
also.

But please remember the original assertions by the gurus. There
is ZERO phase shift through an inductor. There is ZERO amplitude
change through an inductor. This can easily be proven by observing
the lumped inductances in EZNEC. W7EL shot down those arguments
by installing the helix feature in EZNEC. :-)

Never blame malice when ignorance will suffice.

If this person has to confess between ignorance and malicious
behavior, I am sure he would go to jail rather than admit
any ignorance.
--
73, Cecil http://www.qsl.net/w5dxp

John Popelish wrote:
Can you use this example, with varying frequency to explore your
assertion that the time delay (frequency times phase shift) of the coil
varies little over a significant range of frequencies up to self
resonance, and that that delay is about 1/4 cycle of the self resonant
frequency?

Please don't put words in my mouth. What I have previously said
is that the delay can be *ROUGHLY* calculated using the self-
resonant frequency. I said something about +/- 50% accuracy.
Here's what EZNEC reports as the phase shift through the coil
in the traveling wave antenna previously tested at 5.89 MHz.

5.5 MHz: 14.1 deg, 5.89 MHz: 15.7 deg, 6 MHz: 16.2 deg,
7 MHz: 21.4 deg, 8 MHz: 29.5 deg, 9 MHz: 45.9 deg,
10 MHz: 89 deg, 11 MHz: 141.4 deg, 12 MHz: 163.0 deg,
13 MHz: 172.3 deg, 13.7 MHz: 183.82 deg.

The linear delay calculation is off by 59%, not too far from
my 50% rough estimate. Please note that the above values of
delays reported by EZNEC are nowhere near the 3 nS measured
by W8JI in the standing wave environment.
--
73, Cecil http://www.qsl.net/w5dxp

On Sun, 26 Mar 2006 17:21:17 GMT, Cecil Moore
wrote:

I said something about +/- 50% accuracy.
The linear delay calculation is off by 59%, not too far from
my 50% rough estimate.

error is growing faster than the national debt. ;-)

nowhere near the 3 nS measured
by W8JI in the standing wave environment.

On Sun, 26 Mar 2006 16:39:57 GMT, Cecil Moore
wrote:
delay of 7.4 nS

Hmm, giving Tom the same grace of 59% reveals that the figures above,
7.4nS ±59% (4.4 - 11.77)
and
3nS ±59% (1.77 - 4.77)
overlap.

The thing about error (especially when it is in a growth mode
indicating loss of control over the experiment) is that you don't know
where within the band of possible values that the actual value
resides.

So, comparing the one to the other, making a claim that the other is
invalid, must necessarily invalidate both as they are convergent. Such
is the legacy of poor quality control.

It might be tempting to perform a Hail Mary save, by suddenly
declaring they are both right. :-)
but at 59% error, we can all agree that's a fantasy. Stretching your
tolerance for error to fit your argument can lead to any conclusion.