Reg, G4FGQ wrote:
"It may be possible but where is it?"

In the shortwave broadcast plant I worked in 50 years ago we had a 3.5
KW AM Raytheon "Autotune" transmitter we used to talk back to our
program relay transmitting station in another country. We called it our
"order-wire " transmitter. It or its twin were sometimes used for
broadcasting but it was low in power for that job.

This autotune transmitter had a rotary telephone dial on its panel for
programming its mode, operating drequency, etc. You could instruct it to
listen to instructions, then dial in A-3 for AM, followed by the
frequency you wanted it to operate on, such as 15,925, hit the go
button, then stand back and watch the knobs spin as it tuned itself up
completely. including putting the desired power into a dummy load. A
ready lamp informed you it was good to go on the air at the push of a
button. It worked like a charm.

Collins made autotune transmitters which are now military relics of
WW-2. I never toyed with one of those.

Best regards, Richard Harrison, KB5WZI

Reg Edwards wrote:
Not fundamentally different - it only means the magloop tuning will
be
more sensitive. The servo will still try to drive the system to
resonance at zero phase angle.
=================================

Ian, what slightly worries me is -

(1) The resistive component of antenna input impedance, as measured at
the input of the small coupling loop, when the main loop is even only
slightly off-resonant, is altogether different from 50 ohms but is not
included in the bridge balancing process. The diameter of the coupling
loop is fixed. Yet magnitude and phase adjustments react upon each
other as is experienced by a human operator with two variable
controls.

(2) The coupling between the two loops is very loose. We are trying to
adjust the main loop exactly to resonance via a means which is very
insensitive to its resonant condition. Direct voltage and current
sampling connections to the main loop itself are impossible.

(3) We can imagine a situation where the impedance phase-angle is zero
at the measuremnt point, and the green LEDs light up, but which does
not correspond to exact resonance in the main loop. And exact
resonance matters with a magloop.

(4) Because the system is trying to reduce a phase angle to zero in
the presence of two unknowns, instability can result. We can imagine
the system continuously hopping about trying to find the zero.

As you can see, I have difficulty in describing what I think happens
circuitwise. But I shall be convinced only when somebody produces
something which WORKS reliably without human intervention.

It may be possible but where is it?

I see your point... but what do you actually tune for, when you do it
manually?



--
73 from Ian G/GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

Ian White wrote
I see your point... but what do you actually tune for, when you do
it manually?

===============================
Frequency is selected by the receiver, not the transmitter.

The transmitter is OFF when I do it manually and I tune for maximum
noise in the receiver. How do YOU do it? smiley

If I can't do it when the transmitter is ON then neither can an
automatic ATU. It would have to be more clever than I am.
----
Regards, Reg, G4FGQ

Reg Edwards wrote:

Ian White wrote
I see your point... but what do you actually tune for, when you do
it manually?

===============================
Frequency is selected by the receiver, not the transmitter.

The transmitter is OFF when I do it manually and I tune for maximum
noise in the receiver. How do YOU do it? smiley

I freely admit, I've never touched the things... just trying to be
helpful :-)

But what do you think "maximum noise" means? You hope it's going to mean
maximum field strength when you come to transmit, but what does that
actually mean in terms of loop tuning conditions?


If I can't do it when the transmitter is ON then neither can an
automatic ATU. It would have to be more clever than I am.

If we're not clever enough to build an automatic ATU for a magloop, it's
a sign that there's something about magloops we still need to know...
not abandon the idea.

First of all, somebody needs to build a phase detector for an existing
manually tuned loop, and see what results it gives.


--
73 from Ian G/GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

On Mon, 10 Oct 2005 08:18:35 +0100, Ian White G/GM3SEK
wrote:

Reg Edwards wrote:

Ian White wrote
I see your point... but what do you actually tune for, when you do
it manually?

===============================
Frequency is selected by the receiver, not the transmitter.

The transmitter is OFF when I do it manually and I tune for maximum
noise in the receiver. How do YOU do it? smiley

I freely admit, I've never touched the things... just trying to be
helpful :-)

But what do you think "maximum noise" means? You hope it's going to mean
maximum field strength when you come to transmit, but what does that
actually mean in terms of loop tuning conditions?


If I can't do it when the transmitter is ON then neither can an
automatic ATU. It would have to be more clever than I am.

If we're not clever enough to build an automatic ATU for a magloop, it's
a sign that there's something about magloops we still need to know...
not abandon the idea.

First of all, somebody needs to build a phase detector for an existing
manually tuned loop, and see what results it gives.

Ian, Reg, I don't for a moment think automation of a loop tuner is
trivial, but I don't think it is impractical with modern processor
control techniques. For a practical solution, it will probably need
some king of position sense from which to derive end limits and
velocity for implementing a second order control loop.

It is probably a lot like autotuning a PA tank (and that was done in
closed loop linear systems decades ago), except that the loop Q is
probably much higher and the tuning range wider. The same problems
will be encountered in finding the true resonance where phase changes
very quickly from a large positive value to a large negative value or
vice versa depending on the direction of tuning.

It would be interesting Reg, just the install an SWR meter (this is
not a windup!) at your magloop, shut your eyes and tune it up on rx as
you describe, then see whether is is close to minimum SWR )which will
probably be at or very near zero reactance) on tx.

To some extent, the tuner algorithms will be simpler than for a two to
n variable tuner (many autotuners vary more than two components, they
may switch in an autotransformer, or change from L to PI or PI-L
configuration).

I am more interested to hear from someone who says it can't be done,
what they tried that didn't work.

A remote autotuner could be just the thing that makes a magloop a very
attractive, small, frequency versatile antenna (well, for those who
don't have the antenna in the shack or very long arms).

Owen
--

"Ian White G/GM3SEK" wrote in
If we're not clever enough to build an automatic ATU for a magloop,
it's
a sign that there's something about magloops we still need to
know...
====================================

We are able to analyse and predict the behaviour of magloops to any
required degree of precision.

What is missing is how both a magnitude-searching and phase-searching
circuit of an automatic tuner works when denied access to the
magnitude-searching component.

When manually adjusting a tuner it is obvious to the operator that the
controls INTERACT with each other. Both variable controls equally
affect both magnitude and phase. That much can be gleaned from
inspection of the circuitry.

For example, in the case of a T-tuner with two variable capacitors,
the operator cannot concentrate on one variable exclusively to the
other. He continually has to swap from one to the other and obtain a
balance by progressively closer approximations whilst keeping his eyes
on the co-called SWR meter.

An automatic tuner manages to complete the operation by varying both
controls simultaneously. But it is obvious from observation of what
the drive motors are doing, and the time taken to do it, that the
circuit is behaving just like a human operator. Occasionally the
motors even have to reverse and try again.

When denied access to either one of the two variable controls, the
automatic tuner doesn't know what to do next and would become lost.

If the desired impedance magnitude is known to be 50 ohms and is
somehow inserted in the circuit, this is of little assistance to how
the circuit behaves because when the main loop is off-resonance the
actual resistive component is miles away from 50 ohms yet the
automatic tuner is obliged to do something about it. But without the
ability to vary the diameter of the coupling loop, as I say, it is
lost.

So we need something different from and more sophisticated than the
conventional automatic tuner with its relatively simple magnitude and
phase-searching abilities.

I'll believe it when I see one which works.

Regarding your question about manual tuning up for maximum noise (or
signal) in the receiver, at the frequency set by the receiver,
reciprocity rules and fortunately, with modern transceivers, one can
bawl into the microphone and answer a CQ call with confidence that it
can be heard.

But Ian, you already know all this. I have the time and I just like
gabbing about it.

I trust you are comfortably settling down in your new country. I have
spent happy years, in bits, working in Scotland. It is a most
civilised place.
----
Yours, Reg, G4FGQ

"Owen Duffy" wrote
Yes, I am aware that is only adjusting the reactance of the load
presented, but yes, I was aware the R component doesn't vary over a
very wide range over the loop's operating frequency range, and that
if
X was tamed, the line losses would be acceptable.

====================================

Line loss on transmit is relatively unimportant. What matters is
accuracy of tuning up an extremely high Q loop on both transmit and
receive.
----
Reg, G4FGQ

"Owen Duffy" wrote
I am more interested to hear from someone who says it can't be done,
what they tried that didn't work.

===================================
Owen,

I cannot but agree with everything you say and am inclined to say it
can't be done automatically with anywhere near to the required degree
of accuracy.

The obvious automatic way of tuning the loop is to use a remotely
controlled signal generator at the other end of the back yard and tune
up for maximum signal on receive. This should not be very difficult to
arrange but it is too uneconomic to be practical.

So I'll stick to manual tuning.

But I hasten to add that I hope this conversation does not deter
anybody from experimenting with magloops because magloops are, by far,
the most power-efficient of all the small, space-saving,
neighbor-friendly antennas.

Try one on the 160, 80 and 40 meter bands. At higher frequencies
there's usually enough space to erect a dipole or an inverted-L which
will perform at least as well. (Ian, you could perhaps do a useful
article on the subject.)

For the basic design, download program MAGLOOP4 from the website
below.
.................................................. ..........
Regards from Reg, G4FGQ
For Free Radio Design Software go to
http://www.btinternet.com/~g4fgq.regp
.................................................. ..........

Reg Edwards wrote:
If I can't do it when the transmitter is ON then neither can an
automatic ATU. It would have to be more clever than I am.

It would be relatively easy to use the SWR meter driving
current from an MFJ-259 to control the ATU motor.
--
73, Cecil http://www.qsl.net/w5dxp

Reg Edwards wrote:
So we need something different from and more sophisticated than the
conventional automatic tuner with its relatively simple magnitude and
phase-searching abilities.

If you remembered what frequency the mag loop had been tuned
to last time, would that alleviate the need to know phase?
The reason I ask is that is how I auto-tune my screwdriver
without knowing phase.
--
73, Cecil http://www.qsl.net/w5dxp