K7ITM wrote:
On Nov 29, 9:11 am, Jim Kelley wrote:
...

Over the range of a few octaves, propagation delay on the other hand
does not vary to any significant extent as a function of frequency.
Ostensibly, it should be equal to sqrt(LC) series L, shunt C.



Actually, Jim, I do expect it to have considerable frequency
dependence. I think you can find info about this in books that
address the design of travelling-wave tubes.

I can't think of an example of an active (or reactive) device which
doesn't have frequency dependent characteristics. To the extent that
indices of refraction are frequency dependent, propagation velocity
does in fact vary with frequency. If it didn't, we wouldn't see
rainbows. Dielectric constants do indeed have a frequency dependence.
But to first order, at radio frequencies, in amateur applications,
for the purposes of this discussion, and in my opinion, the effect is
less than considerable - particularly if we assume the L and C in
sqrt(LC) are correct at the frequency of interest. ;-)

73, Jim AC6XG

On Nov 30, 10:59 am, Jim Kelley wrote:
K7ITM wrote:
On Nov 29, 9:11 am, Jim Kelley wrote:
...

Over the range of a few octaves, propagation delay on the other hand
does not vary to any significant extent as a function of frequency.
Ostensibly, it should be equal to sqrt(LC) series L, shunt C.

Actually, Jim, I do expect it to have considerable frequency
dependence. I think you can find info about this in books that
address the design of travelling-wave tubes.

I can't think of an example of an active (or reactive) device which
doesn't have frequency dependent characteristics. To the extent that
indices of refraction are frequency dependent, propagation velocity
does in fact vary with frequency. If it didn't, we wouldn't see
rainbows. Dielectric constants do indeed have a frequency dependence.
But to first order, at radio frequencies, in amateur applications,
for the purposes of this discussion, and in my opinion, the effect is
less than considerable - particularly if we assume the L and C in
sqrt(LC) are correct at the frequency of interest. ;-)

73, Jim AC6XG

OK, that leaves us with a difference of opinion, or a difference in
what we are describing. There was an article in "RF Design" maybe 15
years ago now by John Mezak, K2RDX, describing a helical transmission
line model for coils. At the time, he offered free software to
execute the calculations (which also, to me, offered a very practical
way to calculate coil parameters like inductance, effective shunt
capacitance, and first parallel and series self resonances). He later
charged a nominal fee for an improved version of the software, which I
have. For the "100 turn, 10 inch long, 2 inch diameter" coil wound
with 15AWG copper wire, using John's program, I see a variation of
about 2:1 in propagation velocity between 1MHz and 20MHz. Since the
first parallel self-resonant frequency is predicted to be around 8MHz,
it's perhaps not fair to look as high as 20MHz, but even between 1MHz
and 4MHz, I see about 25% change in predicted propagation velocity.

You may say that perhaps John messed all that up terribly, but I don't
think so...and there are other places you can find similar results.
There's an excellent inductance calculator on-line at
http://hamwaves.com/antennas/inductance.html, and though the absolute
value of its prediction of propagation velocity is about 5% different
than Mezak's, they both show very nearly the same percentage change
with frequency.

It might be worth having a bit closer look at, Jim. Perhaps it's just
that you're thinking of a different effect than what these two
programs (and the theory behind them) are modelling.

Cheers,
Tom

On 30 Nov, 12:25, K7ITM wrote:
On Nov 30, 10:59 am, Jim Kelley wrote:





K7ITM wrote:
On Nov 29, 9:11 am, Jim Kelley wrote:
...

Over the range of a few octaves, propagation delay on the other hand
does not vary to any significant extent as a function of frequency.
Ostensibly, it should be equal to sqrt(LC) series L, shunt C.

Actually, Jim, I do expect it to have considerable frequency
dependence. I think you can find info about this in books that
address the design of travelling-wave tubes.

I can't think of an example of an active (or reactive) device which
doesn't have frequency dependent characteristics. To the extent that
indices of refraction are frequency dependent, propagation velocity
does in fact vary with frequency. If it didn't, we wouldn't see
rainbows. Dielectric constants do indeed have a frequency dependence.
But to first order, at radio frequencies, in amateur applications,
for the purposes of this discussion, and in my opinion, the effect is
less than considerable - particularly if we assume the L and C in
sqrt(LC) are correct at the frequency of interest. ;-)

73, Jim AC6XG

OK, that leaves us with a difference of opinion, or a difference in
what we are describing. There was an article in "RF Design" maybe 15
years ago now by John Mezak, K2RDX, describing a helical transmission
line model for coils. At the time, he offered free software to
execute the calculations (which also, to me, offered a very practical
way to calculate coil parameters like inductance, effective shunt
capacitance, and first parallel and series self resonances). He later
charged a nominal fee for an improved version of the software, which I
have. For the "100 turn, 10 inch long, 2 inch diameter" coil wound
with 15AWG copper wire, using John's program, I see a variation of
about 2:1 in propagation velocity between 1MHz and 20MHz. Since the
first parallel self-resonant frequency is predicted to be around 8MHz,
it's perhaps not fair to look as high as 20MHz, but even between 1MHz
and 4MHz, I see about 25% change in predicted propagation velocity.

You may say that perhaps John messed all that up terribly, but I don't
think so...and there are other places you can find similar results.
There's an excellent inductance calculator on-line athttp://hamwaves.com/antennas/inductance.html, and though the absolute
value of its prediction of propagation velocity is about 5% different
than Mezak's, they both show very nearly the same percentage change
with frequency.

It might be worth having a bit closer look at, Jim. Perhaps it's just
that you're thinking of a different effect than what these two
programs (and the theory behind them) are modelling.

Cheers,
Tom- Hide quoted text -

- Show quoted text -

Where can I obtain a copy of Johns program?
TIA
Art

On Nov 30, 12:48 pm, art wrote:
On 30 Nov, 12:25, K7ITM wrote:



On Nov 30, 10:59 am, Jim Kelley wrote:

K7ITM wrote:
On Nov 29, 9:11 am, Jim Kelley wrote:
...

Over the range of a few octaves, propagation delay on the other hand
does not vary to any significant extent as a function of frequency.
Ostensibly, it should be equal to sqrt(LC) series L, shunt C.

Actually, Jim, I do expect it to have considerable frequency
dependence. I think you can find info about this in books that
address the design of travelling-wave tubes.

I can't think of an example of an active (or reactive) device which
doesn't have frequency dependent characteristics. To the extent that
indices of refraction are frequency dependent, propagation velocity
does in fact vary with frequency. If it didn't, we wouldn't see
rainbows. Dielectric constants do indeed have a frequency dependence.
But to first order, at radio frequencies, in amateur applications,
for the purposes of this discussion, and in my opinion, the effect is
less than considerable - particularly if we assume the L and C in
sqrt(LC) are correct at the frequency of interest. ;-)

73, Jim AC6XG

OK, that leaves us with a difference of opinion, or a difference in
what we are describing. There was an article in "RF Design" maybe 15
years ago now by John Mezak, K2RDX, describing a helical transmission
line model for coils. At the time, he offered free software to
execute the calculations (which also, to me, offered a very practical
way to calculate coil parameters like inductance, effective shunt
capacitance, and first parallel and series self resonances). He later
charged a nominal fee for an improved version of the software, which I
have. For the "100 turn, 10 inch long, 2 inch diameter" coil wound
with 15AWG copper wire, using John's program, I see a variation of
about 2:1 in propagation velocity between 1MHz and 20MHz. Since the
first parallel self-resonant frequency is predicted to be around 8MHz,
it's perhaps not fair to look as high as 20MHz, but even between 1MHz
and 4MHz, I see about 25% change in predicted propagation velocity.

You may say that perhaps John messed all that up terribly, but I don't
think so...and there are other places you can find similar results.
There's an excellent inductance calculator on-line athttp://hamwaves.com/antennas/inductance.html, and though the absolute
value of its prediction of propagation velocity is about 5% different
than Mezak's, they both show very nearly the same percentage change
with frequency.

It might be worth having a bit closer look at, Jim. Perhaps it's just
that you're thinking of a different effect than what these two
programs (and the theory behind them) are modelling.

Cheers,
Tom- Hide quoted text -

- Show quoted text -

Where can I obtain a copy of Johns program?
TIA
Art

You might start by asking John. I'm sure he's in the QRZ database.

Cheers,
Tom

K7ITM wrote:


OK, that leaves us with a difference of opinion, or a difference in
what we are describing. There was an article in "RF Design" maybe 15
years ago now by John Mezak, K2RDX, describing a helical transmission
line model for coils. At the time, he offered free software to
execute the calculations (which also, to me, offered a very practical
way to calculate coil parameters like inductance, effective shunt
capacitance, and first parallel and series self resonances). He later
charged a nominal fee for an improved version of the software, which I
have. For the "100 turn, 10 inch long, 2 inch diameter" coil wound
with 15AWG copper wire, using John's program, I see a variation of
about 2:1 in propagation velocity between 1MHz and 20MHz. Since the
first parallel self-resonant frequency is predicted to be around 8MHz,
it's perhaps not fair to look as high as 20MHz, but even between 1MHz
and 4MHz, I see about 25% change in predicted propagation velocity.

You may say that perhaps John messed all that up terribly, but I don't
think so...and there are other places you can find similar results.
There's an excellent inductance calculator on-line at
http://hamwaves.com/antennas/inductance.html, and though the absolute
value of its prediction of propagation velocity is about 5% different
than Mezak's, they both show very nearly the same percentage change
with frequency.

It might be worth having a bit closer look at, Jim. Perhaps it's just
that you're thinking of a different effect than what these two
programs (and the theory behind them) are modelling.

Cheers,
Tom

Hi Tom -

I suspect that for a given coil, depending on construction, L and/or C
may vary enough over several ocataves to resolve any apparent
'dispute' between my comments and the results provided by Mr. Mezak's
modelling program. I do not believe these effects are large enough to
be responsible for the differences being reported in phenomenon under
discussion.

I would be interested in knowing the results your program produces for
the 100 turn, 2" diameter, 10" long coil that Cecil is concerned
about, if you wouldn't mind sharing them.

Thanks and 73,

Jim, AC6XG

On Nov 30, 1:03 pm, Jim Kelley wrote:
K7ITM wrote:
OK, that leaves us with a difference of opinion, or a difference in
what we are describing. There was an article in "RF Design" maybe 15
years ago now by John Mezak, K2RDX, describing a helical transmission
line model for coils. At the time, he offered free software to
execute the calculations (which also, to me, offered a very practical
way to calculate coil parameters like inductance, effective shunt
capacitance, and first parallel and series self resonances). He later
charged a nominal fee for an improved version of the software, which I
have. For the "100 turn, 10 inch long, 2 inch diameter" coil wound
with 15AWG copper wire, using John's program, I see a variation of
about 2:1 in propagation velocity between 1MHz and 20MHz. Since the
first parallel self-resonant frequency is predicted to be around 8MHz,
it's perhaps not fair to look as high as 20MHz, but even between 1MHz
and 4MHz, I see about 25% change in predicted propagation velocity.
You may say that perhaps John messed all that up terribly, but I don't
think so...and there are other places you can find similar results.
There's an excellent inductance calculator on-line at
http://hamwaves.com/antennas/inductance.html, and though the absolute
value of its prediction of propagation velocity is about 5% different
than Mezak's, they both show very nearly the same percentage change
with frequency.

It might be worth having a bit closer look at, Jim. Perhaps it's just
that you're thinking of a different effect than what these two
programs (and the theory behind them) are modelling.

Cheers,
Tom

Hi Tom -

I suspect that for a given coil, depending on construction, L and/or C
may vary enough over several ocataves to resolve any apparent
'dispute' between my comments and the results provided by Mr. Mezak's
modelling program. I do not believe these effects are large enough to
be responsible for the differences being reported in phenomenon under
discussion.

I would be interested in knowing the results your program produces for
the 100 turn, 2" diameter, 10" long coil that Cecil is concerned
about, if you wouldn't mind sharing them.

Thanks and 73,

Jim, AC6XG

Hi Jim,

Just go to the website I provided a link for. The results of the
calcs it performs are certainly within typical experimental tolerance
of the results from Mezak's program. But it's just one model, and you
MUST understand the model and what it's trying to accomplish if you're
going to be successful in applying it.

As for the effects being "large enough to be responsible for...," I
think you will find that the explanation there is adequately covered
by people thinking they understand what someone else has described,
and thinking it's at odds with what they have observed, or with their
own theory (which may or may not be flawed in itself). Like I wrote
before, I'm really not much interested in getting mired down in that
same old stuff (once again). I'm having way too much fun actually
building things with coils (and other parts) and getting them to
perform useful functions. I've learned FAR more about coils and the
circuits they're used in over the past year from designing and
building circuits than I have from looking at the same old stuff here
on r.r.a.a. that's never going to get resolved because someone has too
much invested in wanting to be "right."

Cheers,
Tom

K7ITM wrote:
For the "100 turn, 10 inch long, 2 inch diameter" coil wound
with 15AWG copper wire, using John's program, I see a variation of
about 2:1 in propagation velocity between 1MHz and 20MHz.

Now the question becomes, what was that propagation
velocity at 4 MHz? An EXCEL program that I have gives
a VF of around 0.03 for that coil making a 3 ns delay
through it impossible at 4 MHz.
--
73, Cecil http://www.w5dxp.com