Hi Paul

Couldn't quite understand that (word ot two missing?)

See if I can explain better..

A coax line will radiate (if its a transmitting circuit) and receive
signal (as a receiving cct) if it isnt terminated into the (same)
impedence at both ends of the cable that is the same as the cable
itself. In other words the shield wire isnt always an effective screen.
What is also an issue is if the symmetry of the source and load isnt
preserved. ie if you feed a dipole direct with coax the "balanced"
aerial and "unbalanced" coax will add up to an amount of line radiation.
A balun is often used to balance an unbalanced feeder. This can either
be a wire wound transformer or loops of coax. Feeding a vertical 1/4
wave whip is a good example of feeding an unbalanced antenna with
unbalanced cable..

You can probably imagine that this coax radiating/receiving issue is a
big problem when the cable runs through a cars engine compartment.

Would this need a transformer, inductance/capacitance circuit, static
resistor, or what??

Well that depends on the cct inside. Being a wide band device the
impedence will vary greatly and no doubt have a reactive component. I
have no idea what the norm is for FM broadcast receivers, even those
with 75 ohm or 300ohm connectors! I'd suspect a simple L/C matching
circuit would do it but might even opt for a small toroidal balun. I'd
just assume that the Z would be about 300 ohms and create a 4:1 RF
transformer (thats 2:1 turns) for the 75 ohm coax. The balun will tend
to resolve any reactance and balance problem as well.

My personal preference for an antenna would be a single quad loop fed
over the last 1/4 wave with a piece of RG62 93 ohm coax. This will match
the Z of the antenna (about 110 ohms) with the coax. I'd also coil up
about 5 turns of coax into a 3" loop where it feeds the antenna. This
will resolve the balanced/unbalanced line radiation problem. This is of
course more a transmitter config and given that you want to use over a
wide range of freqs the 1/4 matching section would be a mute point.

Bob, could you explain one mystery to me? A normal aerial arrangement,
whether fed by ribbon or coax, is effectively one continuous loop of
wire. DC resistance is at most a few ohms - ie pretty much a dead
short. Impedence is resistance to AC, ie the signal, but the fact that
impedence is present doesn't remove the dead short. So how come
anything gets around the loop at all?!!

You have to think of the antenna as a tuned AC circuit. In fact
completely throw away any thoughts of a DC circuit havning any effect
whatsoever. Same kind of logic as a power transformer not looking like a
short circuit. Think of the antenna as being feed by an instantaneous
voltage that takes a fixed time to get from the feedpoint to where there
is a "short cicruit" in the wire. The signal from one side of the coax
arrives "in phase" with the one from the other side so no current flows
between them. The trick is in the length of the wire in question. For a
folded dipole the impedence looks like about 300 ohms provided the
antenna is tuned to resonance. If you depart from resonance the
conjugate impedence (resistive plus reactive) will always rise. If you
start feeding this antenna at twice the freqency for what it was
designed you end up with a very high impedence (say 10K plus ohms). You
are now "voltage feeding" the thing.

It can and is a little more compex than this and I am not sure I have
done it justice. Try another analogy. If you are pushing a child on a
swing in sync with it, the energy you put in will be effectively used.
If however you push in the exact opposite sense (by timing 180 degrees
out of phase) you hit what look like a short circuit. Think in terms of
instantaneous voltage in different parts of the antenna that change over
time.

And another: Where does this PD occur? Across the receiving element of
the aerial, or within the element relative to earth? Does this mean an
aerial 'loop" works differently from a single long wire?

Really testing my basic theory here! grin

For most applications you would tend to think that the volts would occur
at the antenna feedpoint as an AC value 180 degrees difference from one
side to the other. You can draw the instantaneous current and voltage on
the wire if you like (and this is handy for working out the shape of the
radiated pattern) but probably isnt helpful in your case.

Also think in terms of maximum power transfer with soruce and load Z...
Reactance of course complicates that and the *real* theory involves
looking at capacitive and indictive reactance..

I hope this all means something to you!

Cheers Bob

Bob Bob wrote:
. . .

A coax line will radiate (if its a transmitting circuit) and receive
signal (as a receiving cct) if it isnt terminated into the (same)
impedence at both ends of the cable that is the same as the cable
itself. In other words the shield wire isnt always an effective screen.

This is a common misconception, but it's not true -- the presence of
current on the outside of a coax line (which is the cause of radiation
and signal pickup) has nothing to do with the impedance match at either
end of the line.

What is also an issue is if the symmetry of the source and load isnt
preserved. ie if you feed a dipole direct with coax the "balanced"
aerial and "unbalanced" coax will add up to an amount of line radiation.

That's not entirely correct, either. A coax line can be just as
"balanced" as twinlead, and twinlead can be just as "unbalanced" as
coax. Imbalance is caused by other factors besides physical symmetry --
see "Baluns: What They Do and How They Do It" in the ARRL Antenna
Compendium, Vol. 2, for example.

. . .


You have to think of the antenna as a tuned AC circuit. In fact
completely throw away any thoughts of a DC circuit havning any effect
whatsoever. Same kind of logic as a power transformer not looking like a
short circuit. Think of the antenna as being feed by an instantaneous
voltage that takes a fixed time to get from the feedpoint to where there
is a "short cicruit" in the wire. The signal from one side of the coax
arrives "in phase" with the one from the other side so no current flows
between them. The trick is in the length of the wire in question. . .

I don't quite follow this, so I might be misinterpreting what was said,
but it doesn't sound quite right. A small loop has a small value of
resistance but a moderate amount of inductance. So the impedance
(composed of the resistance and inductive reactance), although low, is
considerably higher than just the low-frequency value. Although the
reactance provides the lion's share of the impedance, even the
resistance is quite a bit higher than the low frequency value due to
skin effect. As the antenna gets larger, the inductance increases and,
due to capacitive effects, the antenna eventually becomes resonant.
That's the frequency at which the inductive and capacitive reactances
just cancel. A short dipole antenna, on the other hand, has a much lower
impedance at RF than its low frequency value, primarily due to the
capacitance between the two conductors. Its impedance is composed of a
small resistance and a large capacitive reactance.

The current on the center conductor of the coax isn't in phase with the
current on the shield -- it's exactly out of phase with the current on
the inside of the shield. So the same amount of current that flows out
of the inner conductor flows back along the inside of the shield. In
this way, an antenna behaves just like any other electric circuit.

. . .

Roy Lewallen, W7EL

Yes Bob, what you're saying does make some sense and will make even
more after I've read it a few times. So does Roy's.

I've got a pretty good handle on the basics, having a physics degree
and a long interest in hobby electronics. But as black arts go, the
workings of antennae and tuned circuits come pretty high on the list!

Some experimentation is obviously needed. I'll start with a simple
dipole and a run of coax, the outer screen attached to chassis. How
does that sound? Is it worth buying a proper aerial or, being in the
loft, would a home-made rig suffice? What would you suggest
design-wise?

Money isn't really an issue - just that there's no point in buying kit
that turns out to be unnecessary. As I said, the signal is already
improved by hooking up the flying aerial lead to the wire in the duct.

Meanwhile I'll let you guys fight out the theory!

Many thanks for your advice

Paul

Hi Roy

Mmmmm, yet another base belief falls screaming from the sky! grin

It would be interesting to know how it happened (the misconception that
is). The low level thinking (for me) on a transmission line is that for
a pair of wires, the signal on one side is always 180 degrees out of
phase with the other. Any noise induced by a magnetic field would be in
the same phase hence common mode rejection means it isnt "seen" by a RX.
I cant help thinking that unbalanced line is somewhat asymetrical in
that the diameter of the outer conductor would *somehow* have a non
trivial width when it comes to the wavelength of the frequency in use.
Needless to say I havent gone to the extent of plotting magnetic lines
and rereading my base theory stuff. Some things one just has to accept!

I get the *feeling* that balanced feeder is less likely to radiate than
unbalanced.

Why does one use triax in some situations? (cable damage and inadvertent
DC supply grounding aside)

One hopes it is a fair statement to say that any inbalance in the
current on either side of the transmission line (or phase shift ne 180
deg) will result in line radiation.. (However one gets it)

I have grabbed the ARRL balun PDF from the Eznec site and will see how I
go... Have to pay some bills first!

I was trying to explain why an antenna (folded dipole or qaud) isnt
really a short circuit at the operating frequency, without referring to
terms like Xc, Xl, and resonance. I see I failed! What I was trying to
portray was that for the amount of time and distance (aka wavelength)
that the electrons had to travel around the antenna any instant in time
and at any specifc place would never see a short circuit. Or of you like
AC is very different to DC.

What I really want to know is whether Paul's FM RX is now working!

Cheers Bob


Roy Lewallen wrote:
-----snips----

You have to think of the antenna as a tuned AC circuit. In fact
completely throw away any thoughts of a DC circuit havning any effect
whatsoever. Same kind of logic as a power transformer not looking like
a short circuit. Think of the antenna as being feed by an
instantaneous voltage that takes a fixed time to get from the
feedpoint to where there is a "short cicruit" in the wire. The signal
from one side of the coax arrives "in phase" with the one from the
other side so no current flows between them. The trick is in the
length of the wire in question. . .


I don't quite follow this, so I might be misinterpreting what was said,
but it doesn't sound quite right.

Bob Bob wrote:
Hi Roy

Mmmmm, yet another base belief falls screaming from the sky! grin

It would be interesting to know how it happened (the misconception that
is). The low level thinking (for me) on a transmission line is that for
a pair of wires, the signal on one side is always 180 degrees out of
phase with the other.

And why would that be? Unfortunately, there's nothing that forces this
to be true. There *is* something that forces it to be true *inside* a
coaxial cable -- the fact that the shield confines the fields created by
the wires. But then there's the outside to worry about. And there's no
such restriction on a twinlead transmission line.

I won't even try to deal with votages on a transmission line, because
you lose any hope of a "ground" reference when you get more than a
fraction of a wavelength away from the Earth. As far as the currents go,
though, there's no reason at all that the currents on the two wires of a
transmission line can't have any possible amplitude and phase
relationship. These are often separated into "common mode" and
"differential mode" or "odd mode" and "even mode" components for
analytical and mathematical convenience. But there are, at the end of
the day, two different currents on the two wires. In a coaxial cable,
the common and differential mode currents are actually physically
separated, the the former being entirely on the outside and the latter
on the inside. If you consider the sum of the current on the outside of
the shield and the current on the inside of the shield to be the "shield
current", you have exactly the same situtation with coax as for twinlead
-- two conductors with two currents that can have any possible
relationship.

Any noise induced by a magnetic field would be in
the same phase hence common mode rejection means it isnt "seen" by a RX.

True only if the receiver inherently has common mode rejection, which
most don't. Take a look at fig. 4 in the balun article. Imagine the
"output stage" being turned around so it's the input stage of a
receiver. Then suppose a noise source comes along and adds 1 mA downward
to both conductors of the cable, which would represent a perfectly
common mode pickup. The radio connector center conductor current would
increase by 1 mA -- which would be delivered directly to the the
receiver. The current on the inside of the radio chassis would increase
by 1 mA, and the current on the outside of the rig would increase by 2
mA (which, not coincidentally, is the total current resulting from the
noise source). The receiver would have a larger input signal as a result
of that noise.

I cant help thinking that unbalanced line is somewhat asymetrical in
that the diameter of the outer conductor would *somehow* have a non
trivial width when it comes to the wavelength of the frequency in use.
Needless to say I havent gone to the extent of plotting magnetic lines
and rereading my base theory stuff. Some things one just has to accept!

I get the *feeling* that balanced feeder is less likely to radiate than
unbalanced.

This becomes a factor only when the dimensions of the transmission line
become a significant part of a wavelength. At that point, you have other
problems maintaining transmission line operation, and up to that point,
other factors are much more important.


Why does one use triax in some situations? (cable damage and inadvertent
DC supply grounding aside)

There are situations in metrology and low level signal work where noise
rejection on the order of 100 dB and better is required. Those require
special shielding and techniques, but only after exceptional balance has
been achieved by other means. The problems most amateurs have with
current imbalance is orders of magnitude greater, and adding another
shield won't help.

You can put a shield around a twinlead line which will guarantee that
the sum of the currents on the two twinlead conductors and the inside of
the shield is zero. But there can be a separate current on the outside,
just like ordinary coax, which can radiate. You don't generally cure
balance problems by adding a shield.


One hopes it is a fair statement to say that any inbalance in the
current on either side of the transmission line (or phase shift ne 180
deg) will result in line radiation.. (However one gets it)

Yes. That's absolutely correct. A line(*), either twinlead or coax, with
equal magnitude and oppositely phased currents won't radiate
significantly. If the currents have any other relationship, they will.
The common mode current is often defined as (I1 + I2)/2, where I1 and I2
are the currents on the two conductors defined as positive in the same
physical direction. The radiation will be the same as if you had two
parallel conductors each carrying this amount of current, both in phase
and in the same direction -- or a single conductor carrrying (I1 + I2)
and no other current. If the currents are equal in magnitude and
opposite in phase, the common mode current is zero and there's no radiation.

(*) The line is assumed to have spacing very small in terms of wavelength.


I have grabbed the ARRL balun PDF from the Eznec site and will see how I
go... Have to pay some bills first!

I was trying to explain why an antenna (folded dipole or qaud) isnt
really a short circuit at the operating frequency, without referring to
terms like Xc, Xl, and resonance. I see I failed! What I was trying to
portray was that for the amount of time and distance (aka wavelength)
that the electrons had to travel around the antenna any instant in time
and at any specifc place would never see a short circuit. Or of you like
AC is very different to DC.

The problem with extremely simplistic analogies is that a thoughtful
person is bound to try and extrapolate to gain further understanding --
just as you're doing here. And the simpler the analogy, the less he can
extrapolate before the analogy falls apart. I don't think any
explanation of AC phenomena can get very far at all without at least a
rudimentary understanding of capacitance and inductance. If there is,
you'd have to be very careful in applying it.


What I really want to know is whether Paul's FM RX is now working!

Cheers Bob

Bob bob wrote:
"The low level thinking (for me) on a transmission line is that for a
pair of wires, the signal on one side is always 180 degrees out of phase
with the other."

That follows for me from the action of the simplest phase inverter, the
center-tapped coil. Feed one end against the center. The other end is
out-of-phase.

The current in opposite wires of a transmission line is flowing in
opposite directions. Hence, the wires are out of phase.

I think a parallel-wire transmission line has a self-balancing tendency.

Two parallel wires form the simplest transmission line, with the power
source at one end and a load at the other.

Distributed along the wires are series resistance, series inductance,
shunt capacitance, and shunt conductance.

Power travels from the source toward the load in the incident wave.
Velocity of the wave and voltage to current ratio in the wires depend on
construction of the line.

The destributed inductance in one parallel wire couples with its mate to
form a mutual reactance. The two wires are inductively coupled. They`re
also capacitively coupled.

With inductive coupling, Lenz`s law says, "the induced current is in
such direction that it opposes the change that produced it". It`s
reactive, pushing back at the imposed current. It flows in the opposite
direction. This is evident in self induction and secondary currents.
Induced current is out of phase.

With current in one wire inducing our of phase current in the opposite
wire of a long transmission line of closely coupled wires, balance
between the wires is enhanced.

Best regards, Richard Harrison, KB5WZI