Current through coils
 

John Popelish wrote:
wrote:
(snip)
Time delay measurements of current at each terminal of a "bug-catcher
style" loading coil are now at:
http://www.w8ji.com/inductor_current_time_delay.htm

Thank you for posting the test results. But I see no information that
would allow me to reproduce it. What test equipment and what
measurement set-up was used to produce these results?

I used a HP8753C network analyzer with small current transformers
similar to those used in directional couplers.

I calibrated using normal proceedures, and verified calibration by
inserting known transmission lines. For example when I substituted a
very short jumper, time delay was a few picoseconds. When I connected a
10 foot RG-8X jumper, time delay was about 13.5 nS.

For my phase angle measurements I used a dual channel HP vector
voltmeter with a HP generator, and similar current transformers when
measuring current.

I suppose most people would want to use a scope, but it would not be
near the accuracy of a dual channel vector voltmeter or especially a
vector network analyzer.

I have regular test fixture built on blank PC boards, since I do this
stuff every week for work. It does not "fit" a large coil well, so I
had to support the coil on two tall blocks of styrofoam and clip lead
to it.

I do have a large fixture that is a four foot PC board "box" with
various test jacks for connections to probes I use with larger
components, but my bench is to cluttered to fit it right now. In any
event a groundplane several inches away doesn't seem to bother things.
The only thing that moved when I moved the inductor close to the
fixture was the self-resonant frequency came down a few MHz. Time
delays did not change much unless I added an extra foot of clip lead,
then they increased about 1nS.

If it's useful, I guess I could add some more stuff. But probably
nothing extensive until after Dayton.

Since all of my data agrees with data made a few years ago by a
different person using a different method with different equipment, and
since it agrees with reference material I have, I don't see any reason
to treat it like cutting edge results. The physics is pretty solid, and
the measurements agree.

73 Tom

Current through coils
 

wrote:
When I connected a
10 foot RG-8X jumper, time delay was about 13.5 nS.

Let's take a look at the measurement results. That
13.5 nS delay through the coax would make that piece
of RG-8X 1/4WL self-resonant at ~18.5 MHz, higher than
the specified 16 MHz self-resonant frequency for the
coil. So the laws of physics would dictate that the
delay through the coil cannot be less than the delay
through that piece of coax.

By definition, the physical meaning of that piece of
coax being 1/4WL self-resonant at 18.5 MHz is that
it takes 1/2 of a cycle in time for the forward wave
to make a round trip to the end of the coax and back.
1/2WL of a cycle at 18.5 MHz is 27 nS. So the one-
way delay through the coax is 1/2 of 27 or 13.5 nS.

By definition, the physical meaning of that 10" coil
being 1/4WL self-resonant at 16 MHz is that it takes
1/2 of a cycle in time for the forward wave to make
a round trip to the end of the coil and back. 1/2
of a cycle at 16 MHz is 31 nS. So the one-way
delay through that coil is 1/2 of 31 or 16.5 nS.

The 1/4WL self-resonance point *IS* a measure of the
delay through the coil just as it is a measure of
the delay through a piece of transmission line.

If the coil is indeed 1/4WL self-resonant at 16
MHz, the one-way delay through the coil is *already
known* to be 16.5 nS and that is what should have
been measured. The fact that the *known value* of
the delay through the coil was not measured runs up
a red flag and is technical proof that something was
amiss with the reported results.
--
73, Cecil http://www.qsl.net/w5dxp

Current through coils
 

Current through coils
 

John Popelish wrote:
Thanks much. This helps me to visualize your method in a much more
complete way. I think a photo of the test apparatus would make a fine
addition to your web page documenting this result. I am especially
interested on how all this stuff was arrayed in space during the test.

John, would you agree or disagree with me that for
a well-designed coil, the delay through the coil is
fixed by the laws of physics as 1/4WL on the self-
resonant frequency?

If the self-resonant frequency of a well-designed coil
is measured at 16 MHz, then the delay through the coil
is 90 degrees at 16 MHz and therefore equal to 15.625 nS.

Using the self-resonant frequency to determine the
delay is an easy and accurate way to measure that
delay.

If the delay through the coil, measured at 1/4 the
self-resonant frequency, is appreciably different
from the 15.6 nS measured at 16m, then the measurement
contains an error.

Agree/disagree?
--
73, Cecil http://www.qsl.net/w5dxp

Current through coils
 

wrote:
Since all of my data agrees with data made a few years ago by a
different person using a different method with different equipment, and
since it agrees with reference material I have, I don't see any reason
to treat it like cutting edge results.

You've probably hit the nail right on the head there,
Tom. If your results agree with Roy's then you were
again more than likely measuring standing-wave current
and therefore gained nothing by making those measurements.

How do you explain a well-designed coil exhibiting a
measured delay of 15.6 nS at 16 MHz and a measured
delay of 3 nS at 4 MHz? Don't you realize that is an
impossibility according to the laws of physics? If
the coil is well-designed at 16 MHz, it would also
be well-designed at 4 MHz and exhibit very close
to the same delay at both of those frequencies.
Do you really think the delay changed by 81% between
those two frequencies?

Build yourself an SWR meter calibrated for the Z0
of that coil. Measure the SWR in your coil circuit.
I'll bet it will be nearly infinite. Seems you are
guilty of presupposing the proof again.
--
73, Cecil http://www.qsl.net/w5dxp

Current through coils
 

Cecil Moore wrote:
John Popelish wrote:

Thanks much. This helps me to visualize your method in a much more
complete way. I think a photo of the test apparatus would make a fine
addition to your web page documenting this result. I am especially
interested on how all this stuff was arrayed in space during the test.


John, would you agree or disagree with me that for
a well-designed coil, the delay through the coil is
fixed by the laws of physics as 1/4WL on the self-
resonant frequency?

I haven't formed a strong opinion either way, yet. I am not an
inductor expert, but am learning lots of interesting things, here. I
know that filters with sharp resonances (i.e. multiple db ripple
chebychev) often have a delay that varies dramatically over rather
narrow frequency ranges near cut off.

But I am finding the discussion and test results fascinating. I don't
care who has what opinion. I am just interested in what is factual,
and sometimes that can be tricky to determine and understand in a more
general context. That is why I am interested in the details of the
measurement method as much as I am in the result. I am still having a
bit of trouble visualizing how the coil was instrumented and
terminated to get this result. I am also a beginner when it comes to
S parameters.

My reservation with you and few others is your emotional investment in
being correct. It makes your opinions less trustworthy. I get the
feeling that some of you care more about having had the right answer
than what the result tells us about reality.

I have been wrong lots of times, and I got over it (sometimes with
difficulty). I accept that I will be wrong about lots more things
before I die. The best I can hope for is to realize my mistakes as
rapidly and gracefully as possible.

If the self-resonant frequency of a well-designed coil
is measured at 16 MHz, then the delay through the coil
is 90 degrees at 16 MHz and therefore equal to 15.625 nS.

Using the self-resonant frequency to determine the
delay is an easy and accurate way to measure that
delay.

For a pure delay process, like a classical transmission line, or
acoustic delay line, I agree. I am not so sure for something with
more ways energy communicates across it, like an extended inductor.

That is the open question, in my mind.

If the delay through the coil, measured at 1/4 the
self-resonant frequency, is appreciably different
from the 15.6 nS measured at 16m, then the measurement
contains an error.

Agree/disagree?

No. Not yet. When the test method has been agreed upon, and exactly
what that method measures is understood by all interested parties,
there will be no need for such opinions. The results will be the
results. Then we can work on our opinions of what the results mean in
a more general context (how we extrapolate to other, related, but
different cases.

Current through coils
 

John Popelish wrote:
I am still having a
bit of trouble visualizing how the coil was instrumented and terminated
to get this result. I am also a beginner when it comes to S parameters.

I think Tom did what I did the other night. I hooked the coil
across my IC-756PRO's output, used minimum power, and tried
to supply 4 MHz power to the 4+j1250 ohm coil that I have.
It naturally rejected (reflected) virtually all of that power.
I found, as Tom did, that the standing wave current at both
ends has virtually identical phases but that is already known.
The delay through the coil simply cannot be tested in that
test arrangement. Tom just repeated Roy's experiment of a
few years ago and obtained the same meaningless results.
So did I so I didn't even bother to report them.

My reservation with you and few others is your emotional
investment in being correct. It makes your opinions less
trustworthy.

Whoa there, I just made a mental blunder about radiation
resistance and readily admitted it. My emotional investment
is in fighting falsehoods, myths, and old wives' tales.
That's all.

The test method for determining the delay through a piece
of transmission line or a coil is the same as it has been
for more than a century. Find the 1/4WL self-resonant
point and calculate the delay. Other methods, resulting
in far different results, are obviously invalid.
--
73, Cecil http://www.qsl.net/w5dxp

Current through coils
 

Cecil Moore wrote:
John Popelish wrote:

I am still having a bit of trouble visualizing how the coil was
instrumented and terminated to get this result. I am also a beginner
when it comes to S parameters.


I think Tom did what I did the other night. I hooked the coil
across my IC-756PRO's output, used minimum power, and tried
to supply 4 MHz power to the 4+j1250 ohm coil that I have.
It naturally rejected (reflected) virtually all of that power.
I found, as Tom did, that the standing wave current at both
ends has virtually identical phases but that is already known.
The delay through the coil simply cannot be tested in that
test arrangement. Tom just repeated Roy's experiment of a
few years ago and obtained the same meaningless results.
So did I so I didn't even bother to report them.

My reservation with you and few others is your emotional
investment in being correct. It makes your opinions less
trustworthy.

Whoa there, I just made a mental blunder about radiation
resistance and readily admitted it. My emotional investment
is in fighting falsehoods, myths, and old wives' tales.
That's all.

But the goal of such "fights" should be altering other's opinions.
How's that been working out for you? ;-)

The test method for determining the delay through a piece
of transmission line or a coil is the same as it has been
for more than a century.

Have you got a reference to a Bureau of Standards bulletin on this
method to measure inductive current delay? It doesn't work for
filters made of lumped inductors, capacitors and resistors.
Otherwise, there would not be special designs that sacrifice other
properties, to keep delay almost constant as frequency changes.

Wait a second, an inductor at resonance is a filter made of
inductance, capacitance and resistance (and transmission line
effects). Hmm.

Find the 1/4WL self-resonant
point and calculate the delay. Other methods, resulting
in far different results, are obviously invalid.

"Obvious" must be something in the eye of the beholder.

If a 2 port device (Are there really any perfect 2 port devices that
don't have an implied 3rd port?) involves only a single energy
transport mechanism from one port to the other, this is a bit closer
to obvious. But if the device uses competing, parallel energy
transport mechanisms (EM waves, inter turn capacitance, mutual
inductance, etc.) it is less clear that the combination of energy
transport effects has a constant delay effect on a current wave as
frequency changes.

Current through coils
 

John Popelish wrote:
But the goal of such "fights" should be altering other's opinions. How's
that been working out for you? ;-)

That's not my goal at all, John. My goal is to discuss the
technical facts. I really don't care if anyone "alters
their opinions" or not. That has been a personality
characteristic since my early days. My sister just
remarked on that same fact a few days ago. She said,
"You have never cared what other people think about
you." It wasn't a criticism, just an observation.

We may understand the results of Tom's latest
measurement by considering the following:

50 ohm source===1 WL 50 ohm lossless coax===8+j2500 load

Since the transmission line is lossless, this doesn't
change anything except for the additional one cycle delay
through the line.

What's the system SWR? I get 16000:1. I asked Tom
to measure the currents in the absence of a high SWR
and he takes his measurements in a 16000:1 SWR
environment. How well do you think he honored my
request for an SWR of 1:1? Shucks, he only missed
it by 1,600,000%. :-)

There is essentially no net energy flow in the above
network. Why are we suprised to measure equal standing
wave currents on each side of the coil? It wouldn't have
surprised me if Tom had measured *zero* phase shift just
like the lumped-circuit model predicts. The traveling-
wave delay through a coil simply cannot be measured
using Tom's methods.
--
73, Cecil http://www.qsl.net/w5dxp

Current through coils
 

Richard Clark wrote:
On Sun, 12 Mar 2006 02:29:51 GMT, Cecil Moore wrote:


Tom Donaly wrote:

Cecil, have you ever read the book _Don Quixote_, by Cervantes?
There's a character in there you remind me of.

Tom, please don't tell me that you also believe that a distributed-
network analysis using wave reflection theory is "gobbledygook".


Hmm, Tom, let me guess - Dulcinea. The object of Quixote's attention
who never appears, but is always dreamt about.

Hi Richard,
that sounds like Cecil's theory, which he's always ready to
defend with his strong right arm. I can't believe all the fuss he's made
over something as trivial as a loading coil.
73,
Tom Donaly, KA6RUH