In article ,
Wes Stewart wrote:

Bingo. Didn't seem like an "Electrical feature" to me :-)

Yeah. I'm not sure just why there would be substantial interaction if
the two points are connected to the same test point, since the rated
impedance is pretty high even without 10:1 isolators.

But I still submit that when you separate the bridge, insert a DC and
some cabling, you lose the symmetry and the signal measured by the A
probe is not necessarily the same as the signal incident at the input
to the DC. Close maybe, but not something I would rely on.

The signal on the other side of the T-and-attenuator setup wouldn't be
the same as the signal at the input to the DC, certainly, since the
signal at the input of the DC would be affected by the reflected
signal. I don't disagree with you there.

What I suggest, though, is that the signal on the "A" probe (at the
other side of the T from the DC), and a signal as seen at the output
of the DC's "forward" coupler line, ought to be very closely
correlated. They'd differ by the coupler's coupling factor, of
course, and there's be a bit of phase shift from the coupler
(dependent on the coupler line length and the frequency). However,
the loading at the coupler output from the load (or the calibration
short) ought not to affect the signal appearing at the 'forward' tap
on the coupler.

Remember, when doing the calibration there is a 100% reflection. This
can have a huge perturbing effect on the incident signal at the
coupler input if the source is not well matched.

Agreed, and I don't suggest that measuring the incident at the coupler
input is a good idea.


That's why I
originally suggested a pad right at the coupler input, especially if
there is some cabling between the generator (or power splitting tee)
and the DC.

Agreed.

No. The B probe, in the single directional coupler arrangement, is
not measuring -incident-, but reflected signal.

True. I was assuming a double directional coupler, and asserting that
the "forward" output on the coupler will produce a signal equivalent
(except for scaling and perhaps a tad of phase shift) to a signal
taken from the far side of the splitter-and-pads "T".

In any event, Dan has stated that he doesn't have all of this stuff
and is stuck using the DC only. My suggestion holds, put a pad at the
DC input, measure the incident at the DC input and of course, the
reflected at the coupled port.

Yes, that should work quite well, and I think it'd give results pretty
much equivalent to [1] a dual directional coupler or [2] the
splitter-and-two-pads isolation arrangement.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

Wes,

I only have copies of a few pages from the app note showing a bi-directional
coupler. Is your reference that the note shows a bridge?

Thanks - Dan

Wes Stewart wrote:
On Sun, 01 Jan 2006 22:10:09 -0000, (Dave Platt)
wrote:


In article ,
Wes Stewart wrote:


The 8405A manual indicates the use of a power divider, and then a pair
of equal-value pads. One side goes to the probe T for the A
(reference) probe and thence to the termination, and the other goes to
the probe T for the B probe and thence to the device-under-test.

Actually, my manual does not show this. Although I have the full kit
of a resistive tee, two 50 ohm "N" sampling tees and appropriate
terminations, I don't believe Dan does.

The manual is quite clear that the A and B probes need to be connected
to points which are isolated from one another.

I've read this someplace, but again my version of the manual (unless
I'm really missing something) doesn't say it.

The BAMA copy mentions it in paragraph 3-14. Later text indicates
that attaching the two probes to a single point is an appropriate way
to set phase-zero.


Bingo. Didn't seem like an "Electrical feature" to me :-)

Nevertheless, the
directional coupler provides the isolation between probes.

I see the issue, and I think I was conflating two different sorts of
measurement regimes.

The splitter/isolator/pad arrangement I was referring to appears on
page 3-3 of the 8405A manual available at BAMA. It's what's
appropriate for doing an in-line test of a transmission line or other
network, where you want to see the effect of the network itself and
can measure (via probe B) at the network's output.

Page 3-4 shows a somewhat similar hookup, which doesn't include the
resistive pads... I presume because the device-under-test (an
amplifier) is assumed to have high isolation as part of its design.


No, that's very similar to figure 11 in AN77-3 that I mentioned below.
Without inserting the device under test per Fig 3-3, but removing the
right hand 50 ohm load and connecting the device there, reflection
measurements can be performed. That's a technique described in
AN77-3. The configuration is that of a Wheatstone bridge as mentioned
earlier.

But I still submit that when you separate the bridge, insert a DC and
some cabling, you lose the symmetry and the signal measured by the A
probe is not necessarily the same as the signal incident at the input
to the DC. Close maybe, but not something I would rely on.

Remember, when doing the calibration there is a 100% reflection. This
can have a huge perturbing effect on the incident signal at the
coupler input if the source is not well matched. That's why I
originally suggested a pad right at the coupler input, especially if
there is some cabling between the generator (or power splitting tee)
and the DC.


Neither of these hookups wouldn't work for measuring an antenna, since
you can't measure at the antenna's output.


Not so, see above.


Instead, using a
directional coupler provides the necessary isolation, and (as you
point out) lets you determine the incident and reflected signals
accurately.


I've got to clarify this a bit if I can...

If you have the full set of parts per figure 11 in AN77-3 and you are
using them as shown, then with equal loads on the two ends, the
circuit is essentially a resistive Wheatstone bridge in balance with
the null detected by the difference between probes A and B.

In this case, the "incident" signal -is- measured by the A probe and
the effects to the source by a changing load are incorporated into the
measurement.

In the case at hand, at least as I imagine it, there is no longer an
nice tidy resistive Wheatstone bridge, but some cabling and a
directional coupler in the mix. In this case, the generator is no
longer the "source", the source is the signal at the input to the
coupler. It is my belief (unless I change my mind later) that a
sample derived from a resistive divider remote from the input to the
directional coupler is not a true measure of the incident signal.

Hmmm. In the general case, I believe you're correct.

I suspect that the setup shown in the 8405A manual sets up a specific
special case, though. The diagrams and text seem to be defining a
case in which:

- there is a physical and electrical symmetry in the T arrangement -
that is, the power splitter is symmetrical, and the pair of
attenuator pads between the splitter and the (A probe tap) and
(device under test) are matched. The manual makes a point of this
issue.

- The pads being used are matched to the system's transmission line
impedance, so that any reflected signal coming back from the
DUT/coupler sees a proper termination by the source (the pad and
signal generator, in this case) and is not re-reflected.

In this particular situation, I believe that the incident signals
reaching the DUT (the input to the coupler, in this case) and the "A"
probe, would be identical... would they not? The proper termination
of the reflected wave will mean that it won't re-reflect off of the
generator and alter the incident wave. The "A" probe signal (off on
its side of the "T") and a signal read out via the incident-wave tap
on the directional coupler ought to be the same, once the coupling
coefficient is taken into account... no?


No. The B probe, in the single directional coupler arrangement, is
not measuring -incident-, but reflected signal.

In any event, Dan has stated that he doesn't have all of this stuff
and is stuck using the DC only. My suggestion holds, put a pad at the
DC input, measure the incident at the DC input and of course, the
reflected at the coupled port.

On Sun, 01 Jan 2006 16:51:37 -0800, dansawyeror
wrote:

Owen,

Yes, you are right. The single coupler doesn't isolate the channels. Putting a
10 dbm attenuator between the tee and the coupler changes the 50 Ohm reading. I
think a dual directional coupler is required. I will have to put this on hold
until that problem is solved.


Dan, it seems to me that you should be able to make measurements with
a single directional coupler (DC)

You could connect your signal generator to the directional coupler via
a 40dB attenuator, and put the chan A probe T on the sig gen end of
the attenuator. This sample should be fairly independent of the
reflection from the unknown load (to the extent of the 40dB
attenuator), and so approximately proportional to the incident wave
alone.

The sample from the DC "reflected" port (properly terminated) is
fairly independent of the incident wave (depending on the F/B ratio of
the coupler) and so is approximately proportional to the reflected
wave alone.

Calibration of the B channel magnitude with a s/c and o/c taken as
rho=1 provides the basis for measurement of Gamma. The angle of Gamma
should be calibrated to 180 and o deg respectively.

BTW, the angle of Gamma for a 50 ohm termination is unimportant if rho
is very small. The angle of Gamma is real important for s/c and o/c
and ought be almost exactly 180 deg difference (if not, you have a
instrument problem).

Following this procedure, if the magnitude of the B channel on the
unknown load measures for example 9.5dB below the B chan magnitude on
a s/c, then the return loss is 9.5dB and the VSWR is 2:1. rho (the
magnitude of Gamma) is 0.333 and you could measure the phase offset
from the o/c angle to determine the angle of Gamma.

Why won't this work?

Owen
--

Owen,

That idea seems to work. I set it up and then looked and the 'interference', the
change in phase based on changing the pad. Zero pad showed several degrees phase
shift from -40 dbm, 10 dbm showed small shift, and 20, 30, and 40 were all about
equal. I decided on 20 dbm as a practical base.


Thanks - Dan


Owen Duffy wrote:
On Sun, 01 Jan 2006 16:51:37 -0800, dansawyeror
wrote:


Owen,

Yes, you are right. The single coupler doesn't isolate the channels. Putting a
10 dbm attenuator between the tee and the coupler changes the 50 Ohm reading. I
think a dual directional coupler is required. I will have to put this on hold
until that problem is solved.



Dan, it seems to me that you should be able to make measurements with
a single directional coupler (DC)

You could connect your signal generator to the directional coupler via
a 40dB attenuator, and put the chan A probe T on the sig gen end of
the attenuator. This sample should be fairly independent of the
reflection from the unknown load (to the extent of the 40dB
attenuator), and so approximately proportional to the incident wave
alone.

The sample from the DC "reflected" port (properly terminated) is
fairly independent of the incident wave (depending on the F/B ratio of
the coupler) and so is approximately proportional to the reflected
wave alone.

Calibration of the B channel magnitude with a s/c and o/c taken as
rho=1 provides the basis for measurement of Gamma. The angle of Gamma
should be calibrated to 180 and o deg respectively.

BTW, the angle of Gamma for a 50 ohm termination is unimportant if rho
is very small. The angle of Gamma is real important for s/c and o/c
and ought be almost exactly 180 deg difference (if not, you have a
instrument problem).

Following this procedure, if the magnitude of the B channel on the
unknown load measures for example 9.5dB below the B chan magnitude on
a s/c, then the return loss is 9.5dB and the VSWR is 2:1. rho (the
magnitude of Gamma) is 0.333 and you could measure the phase offset
from the o/c angle to determine the angle of Gamma.

Why won't this work?

Owen
--

Now that the setup is reading consistently I will 'test' a loaded 2m monopole
over a 1 m**2 ground plane.

Dan

Owen Duffy wrote:
Perhaps this lab document might help you:

http://emclab.concordia.ca/~trueman/...ent_2_2005.pdf
--

On Mon, 02 Jan 2006 13:34:54 -0800, dansawyeror
wrote:

Owen,

That idea seems to work. I set it up and then looked and the 'interference', the
change in phase based on changing the pad. Zero pad showed several degrees phase
shift from -40 dbm, 10 dbm showed small shift, and 20, 30, and 40 were all about
equal. I decided on 20 dbm as a practical base.

This doesn't make sense... are you using "dbm" to mean decibels of
attenuation, usually written "dB".

The units "dBm" are usually written to qualify a power level with
respect to one milliwatt.

The attenuator on your sig gen might be marked in dBm, but that
applies to the combination of the oscillator, possibly its level
meter, and the attenuator as a system.

Using the wrong terms for things is often a result of a concept gap!

A 20dB attenuator will reduce the effect of the reflected component to
about the same level as you would expect from a practical directional
coupler, more attenuation is better if you have the power from the sig
gen and the VVM probe chan can operate at the higher input level.

Owen
--

dansawyeror wrote:
Now that the setup is reading consistently I will 'test' a loaded 2m
monopole over a 1 m**2 ground plane.

I suggest that you start with an unloaded monopole or some very simple
antenna with a well known impedance. (You will of course have to know
and allow for the effect of the finite ground plane.)

You also need to take measures to prevent coupling between the antenna
and the outside of the feedline. The ground plane you mention will help,
but there can still be substantial coupling. Some high impedance ferrite
beads at the feedpoint and another set about a quarter wavelength down
should reduce the coupling to a small value.

Roy Lewallen, W7EL

I proceeded before reading this note. The procedure was to zero the phase meter
on an short and then to test the loaded 2m vertical. The result was +10 dbm
forward (before the 20 dbm pad) and -50 dbm reflected. The coupler measures
about -14 dbm. The total was about -60 dbm, with 34 db of that due to the pad
and coupler. The net is -26 db forward - reflected.

(The phase angle and reflected ware very touchy. It was almost impossible to
adjust by changing frequency. It was easier to 'adjust' it by sitting very still
and moving my arm.)

The antenna is a copy from the ARRL handbook. It is a 4 inch segment, a 1 inch
long by 3/4 inch diameter 5 turn coil, and a 4 inch tip. It is mounted over a 2
foot square aluminum plate. This antenna should have an input impedance less
then 20 Ohms.

How can it measure very close to 50 Ohms? Is there something wrong with this
analysis?

Thanks - Dan

Roy Lewallen wrote:
dansawyeror wrote:

Now that the setup is reading consistently I will 'test' a loaded 2m
monopole over a 1 m**2 ground plane.


I suggest that you start with an unloaded monopole or some very simple
antenna with a well known impedance. (You will of course have to know
and allow for the effect of the finite ground plane.)

You also need to take measures to prevent coupling between the antenna
and the outside of the feedline. The ground plane you mention will help,
but there can still be substantial coupling. Some high impedance ferrite
beads at the feedpoint and another set about a quarter wavelength down
should reduce the coupling to a small value.

Roy Lewallen, W7EL

dansawyeror wrote:
I proceeded before reading this note. The procedure was to zero the
phase meter on an short and then to test the loaded 2m vertical. The
result was +10 dbm forward (before the 20 dbm pad) and -50 dbm
reflected. The coupler measures about -14 dbm. The total was about -60
dbm, with 34 db of that due to the pad and coupler. The net is -26 db
forward - reflected.

(The phase angle and reflected ware very touchy. It was almost
impossible to adjust by changing frequency. It was easier to 'adjust' it
by sitting very still and moving my arm.)

The antenna is a copy from the ARRL handbook. It is a 4 inch segment, a
1 inch long by 3/4 inch diameter 5 turn coil, and a 4 inch tip. It is
mounted over a 2 foot square aluminum plate. This antenna should have an
input impedance less then 20 Ohms.

How did you arrive at this figure? I wouldn't hazard a guess without
modeling it.

How can it measure very close to 50 Ohms?

1. Inductor loss.
2. Effect of finite size ground plane.
3. Coupling to feedline.
4. Measurement error.

Is there something wrong with
this analysis?

I don't know. What should the impedance really be?

Roy Lewallen, W7EL

"Owen Duffy" wrote

Using the wrong terms for things is often a result of a concept gap!

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

.. . . . . and using the wrong name for an SWR meter often results in a
concept gap.
----
Reg.