I would like to know if the follow technique will work to measure resonance of
an antenna. I know have a working HP 8405a. It has been a struggle to repair it
and to align it, however it now appears to be working.

My plan is to connect a bi-directional coupler to the 8405a. The coupler will be
terminated in a 50 Ohm terminator. The phase will be read for this
configuration. The next step is to replace the 50 Ohm terminator with a direct
connection to an unknown antenna. I will then change the frequency until the
phase angle matchs the 50 Ohm phase angle.

Does this make sense?

What are the draw backs?

How close will this method get compared to a noise bridge?

Thanks - Dan kb0qil

I would like to know if the follow technique will work to measure resonance
of an antenna. I know have a working HP 8405a. It has been a struggle to
repair it and to align it, however it now appears to be working.

My plan is to connect a bi-directional coupler to the 8405a. The coupler
will be terminated in a 50 Ohm terminator. The phase will be read for this
configuration. The next step is to replace the 50 Ohm terminator with a
direct connection to an unknown antenna. I will then change the frequency
until the phase angle matchs the 50 Ohm phase angle.

Does this make sense?

What are the draw backs?

How close will this method get compared to a noise bridge?

Thanks - Dan kb0qil

Dan, the phase angle of the reflection coefficient of a matched load is
unimportant, and is likely to be anything. Specifically; if it is a dot on
the Smith Chart center -- who cares? Typically the return loss of a good
quality load could well be 40dB. Again, if you are measuring beyond the
frequency range of the coupler you must determine its directivity. It is
very likely that you are reading the phase of poor directivity and not the
50 ohm load. Also if you are using the 8405A, do you have all the
accessories? In particular the BNC ( + 50 ohm thru) probe attachments which
should be connected directly to your coupler ports. If you require N - BNC
adapters to connect to the coupler, then as long as you have the same
structures on each port, any phase error will be cancelled out. It is also
good practice to attach a high quality 20 dB pad to the input of the coupler
(thus ensuring your source return loss approaches 40 dB) -- you cannot
always be certain that your source, plus coax, provides a good match. Note
if the unit cannot lock with the 20 dB pad, then at least use a 10 dB pad.

Another test is to connect both probes to the same source via a BNC "T"
adapter. Do you show zero phase error?

Interesting you have an 8405A, it is many years since I used one. I see
them priced from $250.00 to a fully refurbished unit from Tucker
(www.tucker.com) for $2,500.00.

Regards,

Frank

On Mon, 05 Dec 2005 21:18:33 -0800, dansawyeror
wrote:

I would like to know if the follow technique will work to measure resonance of
an antenna. I know have a working HP 8405a. It has been a struggle to repair it
and to align it, however it now appears to be working.

My plan is to connect a bi-directional coupler to the 8405a. The coupler will be
terminated in a 50 Ohm terminator. The phase will be read for this
configuration. The next step is to replace the 50 Ohm terminator with a direct
connection to an unknown antenna. I will then change the frequency until the
phase angle matchs the 50 Ohm phase angle.

Does this make sense?

Sort of, but not exactly. As I described in another thread, the
pitfall is the fact that the phase lengths of the coupler are
different. You could set the phase offset to zero at frequency X but
when you change to frequency Y, there is a built-in error.

The correction for this is to use a line stretcher that (more-or-less)
equalizes the phase lengths with respect to frequency.


What are the draw backs?

See above.

How close will this method get compared to a noise bridge?

With a little care you can make measurements that leave noise bridge
measurements in the dust... not even close. With -lots- of care, you
can approach network analyzer accuracy... more than is needed for
practical antenna measurements.

You need to see H-P App note AN77-3. When I get time---and with my
dial-up connection---I mean lots of time, I will try to post a 13
MByte .pdf on qsl.net, if that's working.

On Tue, 06 Dec 2005 07:45:36 -0700, Wes Stewart
wrote:

On Mon, 05 Dec 2005 21:18:33 -0800, dansawyeror
wrote:

I would like to know if the follow technique will work to measure resonance of
an antenna. I know have a working HP 8405a. It has been a struggle to repair it
and to align it, however it now appears to be working.

My plan is to connect a bi-directional coupler to the 8405a. The coupler will be
terminated in a 50 Ohm terminator. The phase will be read for this
configuration. The next step is to replace the 50 Ohm terminator with a direct
connection to an unknown antenna. I will then change the frequency until the
phase angle matchs the 50 Ohm phase angle.

Does this make sense?

Sort of, but not exactly. As I described in another thread, the
pitfall is the fact that the phase lengths of the coupler are
different. You could set the phase offset to zero at frequency X but
when you change to frequency Y, there is a built-in error.

I should have added that you don't do this with a 50 ohm load, but a
short circuit instead. You want a big reflection with a known phase
for a calibration standard.

Sort of, but not exactly. As I described in another thread, the
pitfall is the fact that the phase lengths of the coupler are
different. You could set the phase offset to zero at frequency X but
when you change to frequency Y, there is a built-in error.

I should have added that you don't do this with a 50 ohm load, but a
short circuit instead. You want a big reflection with a known phase
for a calibration standard.

The first step is to establish the accuracy of the test equipment. A
short/open at the output port of the directional coupler should produce a
frequency independant phase shift between the forward and reverse coupled
port. If you are not getting 180/0 deg, then there is something
fundamentally wrong with the test set up. Connecting both probes of the
8405A to the same source should establish if any errors exist in the vector
voltmeter. Using the BNC/probe adapters is essential to maintain
repeatability in your measurements.

Reference HP's 778D coupler at
http://cp.literature.agilent.com/lit.../5952-8133.pdf :

"
Impedance Measurements

The 778D is also well suited for measurements of impedance when used with

the Agilent 8405A vector voltmeter. The technique is described in
Application

Note 77-3, Measurement of Complex Impedance, available at your nearest

Agilent sales office. Again, a reflectometry technique is used. With the
vector

voltmeter, however, both magnitude and phase angle of the reflection
coefficient

can be measured. This setup is shown in Figure 3.

Data can be read from the two meters of the vector voltmeter and transferred

directly to a Smith Chart to provide impedance of such devices as antennas
or

other passive components.


"It is many years since I have seen AN77-3, but I seem unable to find a copy
on the web.

Frank

On Tue, 06 Dec 2005 16:48:32 GMT, "Frank"
wrote:

Sort of, but not exactly. As I described in another thread, the
pitfall is the fact that the phase lengths of the coupler are
different. You could set the phase offset to zero at frequency X but
when you change to frequency Y, there is a built-in error.

I should have added that you don't do this with a 50 ohm load, but a
short circuit instead. You want a big reflection with a known phase
for a calibration standard.

The first step is to establish the accuracy of the test equipment. A
short/open at the output port of the directional coupler should produce a
frequency independant phase shift between the forward and reverse coupled
port.

Maybe I misunderstand, but what you seem to be saying is that if I put
Probe A at point X on a transmission line and Probe B at point X+Y
(Y0), the phase difference with respect to frequency will not change.

The directional coupler is no different, it is sampling the main line
at two physically different locations.


If you are not getting 180/0 deg, then there is something
fundamentally wrong with the test set up. Connecting both probes of the
8405A to the same source should establish if any errors exist in the vector
voltmeter. Using the BNC/probe adapters is essential to maintain
repeatability in your measurements.

Reference HP's 778D coupler at
http://cp.literature.agilent.com/lit.../5952-8133.pdf :

"
Impedance Measurements

The 778D is also well suited for measurements of impedance when used with

the Agilent 8405A vector voltmeter. The technique is described in
Application

Note 77-3, Measurement of Complex Impedance, available at your nearest

Agilent sales office. Again, a reflectometry technique is used. With the
vector

voltmeter, however, both magnitude and phase angle of the reflection
coefficient

can be measured. This setup is shown in Figure 3.

Data can be read from the two meters of the vector voltmeter and transferred

directly to a Smith Chart to provide impedance of such devices as antennas
or

other passive components.


"It is many years since I have seen AN77-3, but I seem unable to find a copy
on the web.

Frank

The pertinent pages are he

http://www.qsl.net/n7ws/Pages%20from%20AN77-3.pdf

On Tue, 06 Dec 2005 11:37:04 -0700, Wes Stewart
wrote:


Maybe I misunderstand, but what you seem to be saying is that if I put
Probe A at point X on a transmission line and Probe B at point X+Y
(Y0), the phase difference with respect to frequency will not change.

The directional coupler is no different, it is sampling the main line
at two physically different locations.

With respect, I think you two guys are talking about different kinds
of couplers.

Wes, your earlier description makes is clear that samples in the
coupler you described are located at equidistant from their nearest
ends, but not in the centre, so at different positions in the coupler.
I think the coupled lines type of coupler for lower frequency use
might fit this category.

Frank assumes a coupler where the samples for both ports are taken at
the same physical location on the main line. I think a crossed
waveguide coupler might fit this category (depending on the way the
coupling holes are implemented).

Owen
--

On Tue, 06 Dec 2005 19:18:50 GMT, Owen Duffy wrote:

On Tue, 06 Dec 2005 11:37:04 -0700, Wes Stewart
wrote:


Maybe I misunderstand, but what you seem to be saying is that if I put
Probe A at point X on a transmission line and Probe B at point X+Y
(Y0), the phase difference with respect to frequency will not change.

The directional coupler is no different, it is sampling the main line
at two physically different locations.

With respect, I think you two guys are talking about different kinds
of couplers.

Wes, your earlier description makes is clear that samples in the
coupler you described are located at equidistant from their nearest
ends, but not in the centre, so at different positions in the coupler.
I think the coupled lines type of coupler for lower frequency use
might fit this category.

Frank assumes a coupler where the samples for both ports are taken at
the same physical location on the main line.

May be, but his link:

http://cp.literature.agilent.com/lit.../5952-8133.pdf

shows exactly what I'm talking about in Figure 3. Note the line
stretcher on one sample port. Because I haven't seen the inside of one
of these I don't know where the coupled arms reside with respect to
each other but the fact that H-P shows the line stretcher tells me
that they must have some (unavoidable, if not purposeful) asymmetry.
They claim 4 degree phase tracking but it's unclear to me whether they
mean the coupled arms with respect to the main line or to each other.


I think a crossed
waveguide coupler might fit this category (depending on the way the
coupling holes are implemented).



Owen

"Owen Duffy" wrote in message
news
On Tue, 06 Dec 2005 11:37:04 -0700, Wes Stewart
wrote:


Maybe I misunderstand, but what you seem to be saying is that if I put
Probe A at point X on a transmission line and Probe B at point X+Y
(Y0), the phase difference with respect to frequency will not change.

The directional coupler is no different, it is sampling the main line
at two physically different locations.

With respect, I think you two guys are talking about different kinds
of couplers.

Wes, your earlier description makes is clear that samples in the
coupler you described are located at equidistant from their nearest
ends, but not in the centre, so at different positions in the coupler.
I think the coupled lines type of coupler for lower frequency use
might fit this category.

Frank assumes a coupler where the samples for both ports are taken at
the same physical location on the main line. I think a crossed
waveguide coupler might fit this category (depending on the way the
coupling holes are implemented).

Owen

The type of coupler I am thinking of is indeed where the coupling lines for
forward and reflected are in the same physical region. The point I was
making refers particularly to the HP 778D; where it is specified in:
http://cp.literature.agilent.com/lit.../5952-8133.pdf
"Data can be read from the two meters of the vector voltmeter and
transferred directly to a Smith Chart". Indicating that there is no
significant phase error over the nominal bandwidth of the coupler. It is
certainly something I have always taken for granted, but in recent years
have been spoiled by Agilent's VNAs. Even Matthaei, Young, and Jones'
classic text does not get into specifics of the phase response of TEM
couplers. If I get a chance in the near future, I will run a phase response
of an HP778D coupler, and/or also run an Eagleware simulation of a co-planar
WG coupler.

Just read Wes' comments on the line stretcher on the above pdf. I must
admit I also wondered about that, but assume it is intended as a cal
adjustment for a short/open standard. Even VNAs, in their non calibrated
state, show a pretty good dot on the right/left of the Smith Chart for an
open/short in the lower hundreds of MHz.

Incidentally I have tried to access the
http://www.qsl.net/n7ws/Pages%20from%20AN77-3.pdf which is unbelievably
slow, but I keep getting error messages saying the file is corrupted. I am
using a 2.5 Mb ADSL, which is normally pretty fast.

Frank

"Wes Stewart" wrote in message
...
On Tue, 06 Dec 2005 19:18:50 GMT, Owen Duffy wrote:

On Tue, 06 Dec 2005 11:37:04 -0700, Wes Stewart
wrote:


Maybe I misunderstand, but what you seem to be saying is that if I put
Probe A at point X on a transmission line and Probe B at point X+Y
(Y0), the phase difference with respect to frequency will not change.

The directional coupler is no different, it is sampling the main line
at two physically different locations.

With respect, I think you two guys are talking about different kinds
of couplers.

...Owen's coupler location discussion...
May be, but his link:

http://cp.literature.agilent.com/lit.../5952-8133.pdf

shows exactly what I'm talking about in Figure 3. Note the line
stretcher on one sample port. Because I haven't seen the inside of one
of these I don't know where the coupled arms reside with respect to
each other but the fact that H-P shows the line stretcher tells me
that they must have some (unavoidable, if not purposeful) asymmetry.
They claim 4 degree phase tracking but it's unclear to me whether they
mean the coupled arms with respect to the main line or to each other.


I have done this type of measurement and the coupler is somewhat
un-important in regard to where on the line its samples are taken from. The
line stretcher does the compensation to place both samples at the same
point on the main line. It is set in a calibration process to get to that
point (seems to mee only a short is required). Memory is starting to fade
here, but if you want the samples to be "at the plane of the DUT" then you
either must use two line stretchers or use the vector voltmeter capability
to compensate for the difference in sample location vs. DUT location. If I
recall correctly, this is simply a phase offset and the old HP vector
voltmeter has such an offset capability.

Look at it this way. You want the FWD and REF phases to be at the DUT. To
do this you must put an equal (electrical) length of line in all three
sections of line - the desired, the FWD sample and the reverse sample. I
admit the reflected part has an intuitive glitch that I can't resolve at
this time since it is always longer, but I know this works. You can sweep
the set-up and the phase of the two samples will sit right there on that of
a short . This rells you that the samples are "at the DUT".

Right??

73, Steve, K,9.D;C'I