On Tue, 27 Sep 2005 00:24:51 +0000 (UTC), "Reg Edwards"
wrote:


You must have been reading what I've been writing for the last 6 or 7
years.

Reg, Yes, I have probably read and learned a good deal from stuff you
have written, but I have skipped over a lot of what you have
written... probably most noticeably when you and others trade kicks to
the groin to see who is the last man standing.

Back on topic:

A point that you hinted at, but might have been overlooked by some is
that it can be relatively unimportant that the SWR bridge's sampling
line has the same characteristic impedance as the impedance at which
its detector has been nulled.

For example, a typical SWR meter designed originally for 75 ohms, with
a 0.1m long ideal 75 ohm sampling section, but with the detector
adjusted to read nil reflected power with a 50+j0 load on the
"antenna" terminals of the meter, will in most cases operate just as
well as a 50 ohm SWR meter on 7MHz, as the detector will truly show
when it has a 50 ohm load, the indicated VSWR for other loads will
substantially correct (ie within typical accuracy for the type of
instrument), and the insertion VSWR (~1.02:1) because of the 0.1m of
75 ohm line will be insignificant in practice.

In many amateur reflectometer designs (and in some commercial
implementations), very little attention has been given to the
characteristic impedance of the sampling section, and in some cases to
the insertion VSWR (that results).

I recall testing a relatively expensive SWR meter rated from 1.8 to
150MHz, and noting that whilst it indicated a VSWR1.1 at 144MHz on a
good dummy load a Bird 43 ahead of it indicated an insertion VSWR
1.5:1. So whilst it was good at indicating a 50+j0 ohm load on its
"antenna" terminals, it was not very capable of delivering that load
to its "transmitter" terminals.

Owen
--

Owen Duffy wrote:
On Tue, 27 Sep 2005 00:24:51 +0000 (UTC), "Reg Edwards"
wrote:


You must have been reading what I've been writing for the last 6 or 7
years.

Reg, Yes, I have probably read and learned a good deal from stuff you
have written, but I have skipped over a lot of what you have
written... probably most noticeably when you and others trade kicks to
the groin to see who is the last man standing.

Back on topic:

A point that you hinted at, but might have been overlooked by some is
that it can be relatively unimportant that the SWR bridge's sampling
line has the same characteristic impedance as the impedance at which
its detector has been nulled.

For example, a typical SWR meter designed originally for 75 ohms, with
a 0.1m long ideal 75 ohm sampling section, but with the detector
adjusted to read nil reflected power with a 50+j0 load on the
"antenna" terminals of the meter, will in most cases operate just as
well as a 50 ohm SWR meter on 7MHz, as the detector will truly show
when it has a 50 ohm load, the indicated VSWR for other loads will
substantially correct (ie within typical accuracy for the type of
instrument), and the insertion VSWR (~1.02:1) because of the 0.1m of
75 ohm line will be insignificant in practice.

In many amateur reflectometer designs (and in some commercial
implementations), very little attention has been given to the
characteristic impedance of the sampling section, and in some cases to
the insertion VSWR (that results).

I recall testing a relatively expensive SWR meter rated from 1.8 to
150MHz, and noting that whilst it indicated a VSWR1.1 at 144MHz on a
good dummy load a Bird 43 ahead of it indicated an insertion VSWR
1.5:1. So whilst it was good at indicating a 50+j0 ohm load on its
"antenna" terminals, it was not very capable of delivering that load
to its "transmitter" terminals.

Owen
--

Inaccurate crap equipment has nothing to do with the arguement.

--
Jim Pennino

Remove .spam.sux to reply.

Owen,

SWR meters with a sampling line.

The only experience I've had has been I once made one for HF. It was
of the type where a second wire is drawn alongside the inner conductor
of a short length of coaxial line of impedance in the same street as
the system it is to work with.

Operating frequencies covered the whole of the HF band. That is a very
wide band. Which indicates that line length plays no part in
measuring accuracy once calibrated.

To explain how the thing works it is necessary to return to what it
really is. It is a resistance bridge. All so-called SWR meters,
whatever the circuit or form of construction, are resistance bridges.

The bridge has 3 internal ratio arms. The 4th arm is the variable
transmitter load. If all 4 arms are of same resistance we have a very
sensitive arrangement suitable for QRP transmitters. However, 3/4 of
the TX power is dissipated in the 3 internal bridge arms.

For higher power transmitters it is necessary to use high ratios for
the ratio arms. In the case of meters which use a little ferrite ring
as a current transformer, a resistor of the order of 30 to 100 ohms
can be shunted across the current transformer secondary winding while
the primary winding has an input resistance of the order of 0.1 ohms
which forms the value of the ratio arm in series with the external
load. This 0.1-ohm arm is capable of carrying the load current of
several amps with only a small power loss.

The other two ratio arms can be a pair of high value resistors in the
same ratio as occurs via the current transformer. If the input
resistance of the current transformer is 0.1 ohms then the bridge
ratio is 50 / 0.1 = 500:1 where 50 ohms is the usual value of the
load resistance when the bridge is balanced and SWR = 1:1

The two high impedance arms can be capacitors in the same ratio of
500:1 which have zero power dissipation but have a minor effect on
accuracy. They introduce a small phase angle into the load as seen by
the transmitter through the meter. The error increases with
increasing frequency.

It will be seen that the take-off point is effectively the same for
both current and voltage.

Returning to the so-called sampling line.

There is a bridge configuration which is not quite so obvious. But
instead of a current transformer the current is picked off by means of
a short length of wire in parallel with the coaxial inner conductor by
virtue of their mutual inductance. The line is too short for
propagation effects to play a significant part.

Voltage is picked off at the same point by virtue of the capacitance
between the wire and coaxial inner conductor. The phase relationship
between volts and amps can be reversed just by reversing the direction
of propagation through the meter.

The bridge ratio is set partially by the ratio of the impedances Zo of
the additional wire and inner coax conductor. The length of coaxial
line affects only the bridge sensitivity and power dissipated in the
meter. As you must be aware, sensitivity falls of fast with decreasing
frequency and 160 meters was my favourite band. So the home-brewed
meter was soon discarded and I returned to ferrite rings.

I was left with the impression it was very easy to make and that
almost anything would work.

Hope you can understand the foregoing.
----
Reg.

Reg Edwards wrote:
There is a bridge configuration which is not quite so obvious. But
instead of a current transformer the current is picked off by means of
a short length of wire in parallel with the coaxial inner conductor by
virtue of their mutual inductance. The line is too short for
propagation effects to play a significant part.

The pickup lines in my Heathkit HM-15 are terminated on one
end with a 50 ohm resistor. One pickup line thus attenuates
the reflected traveling wave and allows the forward traveling
wave to be rectified. The other pickup line attenuates the
forward traveling wave and allows the reflected traveling wave
to be rectified. Knowing the peak values of both of these two
traveling waves allows a calibrated meter to indicate SWR.
--
73, Cecil http://www.qsl.net/w5dxp

Cec, you have YOUR explanation and I have MY explanation.

Which is the most simple?

There is a bridge.

When the variable arm, the load, is 50 ohms the bridge is balanced and
the meter indicates SWR = 1:1

When the variable arm is either 0 ohms or infinite ohms, the meter
indicates SWR = infinity :1

What can be more simple than that? How it works can be visualised.

But the meter is ambiguous. It cannot distinguish between loads of 0
ohms and infinite ohms. Additional information is required.

This serious ambiguity also applies to your weird contraption. ;o)
----
Regards, Reg.

Reg Edwards wrote:
Cec, you have YOUR explanation and I have MY explanation.

Mine is a lot simpler. The Heath HM-15 has two pickup elements.
If you install a Z0 resistor load at one end it "picks up" the forward
wave. If you install a Z0 resistor load at the other end it "picks up"
the reflected wave. The two pickup voltages are rectified and compared
through a calibration procedure.

The parts that came with the HM-15 kit in the 50s-60s included two
72 ohm resistors. RG-ll was very popular at the time. If one wanted
a 72 ohm SWR meter, one installed the 72 ohm resistors. If one wanted
a 50 ohm SWR meter, one installed the 50 ohm resistors. A switch could
be installed that switched between 50 ohms and 72 ohms calibration.

This serious ambiguity also applies to your weird contraption. ;o)

Actually, the Heathkit design concept is easier to understand than
is the bridge explanation or the toroid-pickup/phasor-addition
explanation. The first SWR meter I built in the 50s, used two
lengths of insulated wire shoved under the braid of the coax.
It worked but, at the time, I had no idea why it worked. Heath's
little slotted line pickup device was pretty slick. I sometimes
see them for sale at hamfests.
--
73, Cecil http://www.qsl.net/w5dxp

On Tue, 27 Sep 2005 15:48:35 +0000 (UTC), "Reg Edwards"
wrote:
But the meter is ambiguous. It cannot distinguish between loads of 0
ohms and infinite ohms. Additional information is required.

Hi Reggie,

Without recourse to that "additional information," explain how you
achieve the unambiguous by your method of probing lines (be they
parallel, coaxial, or waveguide).

In other words, your objection is a non sequitur, it is meaningless
because you need the same additional information and you cannot
demonstrate any measurable difference between the manifold methods of
coming to the same determination.

Of course, if you throw a spanner in the other guy's gear-box, you
might win the race.

73's
Richard Clark, KB7QHC

Rich, your abuse of the English language renders it impossible for me
or anybody else to make any sense of what you are waffling about.
----
Punchinello, G4FGQ

On Tue, 27 Sep 2005 19:13:16 +0000 (UTC), "Reg Edwards"
wrote:

Rich, your abuse of the English language renders it impossible for me
or anybody else to make any sense of what you are waffling about.

Aw Reggie,

Are you using a prescription grade wine glass when you were trying to
read it? Or can we blame it on the grape? No ifs ands or buts now
because with each new post the question becomes more remote and harder
for you to answer.

73's
Richard Clark, KB7QHC

Cec, I notice that you and others have begun to use my description of
"indicate" rather than "measure".
----
Reg.