HP 8405 and negative resistance of antenna?
 

Hi, Gentlemen -

I have an HP 8405A Vector Voltmeter. I have the App Note "Measuring Complex
Impedance" by HP (thanks to Wes Stewart). I have made a few measurements
following (maybe) the HP paper. I have probably been sloppy with the setup
and procedure because I came up with a negative resistance a couple of
times.

The HP paper discusses negative resistance but does not mention where one
would encounter it. I assume it would normally be associated with active
device measurements. Is this true? If I measure a (passive) load and find
that the resistance is negative, shouldn't I doubt my setup and/or
procedure? Is there any condition where a passive, real-world, network could
show a negative resistance?

By the way, I have not gotten a negative resistance reading while using a
dual directional coupler. I have gotten the negative resistance reading only
when using the power splitter method. So far.

Thanks for any insight on this.

John KD5YI

"The other John Smith" wrote
Is there any condition where a passive, real-world, network
could show a negative resistance?
____________

"Negative towers' exist in some MW broadcast directional arrays. Here is a
link to a paper about it: http://www.qsl.net/km5kg/negative.htm. Don't
know if that addresses your situation or not...

RF

I see from my records that you have EZNEC. Open EZNEC example file
4square.ez, and click Src Dat. You'll see that element 1 has a negative
resistance. In the main EZNEC window, open the File menu, then click
Edit Current Antenna Notes. This contains an explanation of the
phenomenon. It can happen only when there are multiple sources, since
the power to create the negative resistance at one source comes from
power supplied by others. If you use a feed system to distribute power
to various elements from a single source, the resistance at that source
can never be negative.

Active devices can also exhibit a negative resitance, but again, there
has to be some source of power, since a negative resistance sources
power to whatever is connected to it.

Your result could be due to measurement error, erroneous setup or
procedure, or, if you're measuring an antenna, RF on the antenna from
outside sources. The latter makes antenna impedance measurement
difficult without a sharply tuned detector, which the 8405A doesn't have.

Roy Lewallen, W7EL

The other John Smith wrote:

Hi, Gentlemen -

I have an HP 8405A Vector Voltmeter. I have the App Note "Measuring Complex
Impedance" by HP (thanks to Wes Stewart). I have made a few measurements
following (maybe) the HP paper. I have probably been sloppy with the setup
and procedure because I came up with a negative resistance a couple of
times.

The HP paper discusses negative resistance but does not mention where one
would encounter it. I assume it would normally be associated with active
device measurements. Is this true? If I measure a (passive) load and find
that the resistance is negative, shouldn't I doubt my setup and/or
procedure? Is there any condition where a passive, real-world, network could
show a negative resistance?

By the way, I have not gotten a negative resistance reading while using a
dual directional coupler. I have gotten the negative resistance reading only
when using the power splitter method. So far.

Thanks for any insight on this.

John KD5YI

Active devices can also exhibit a negative resitance, but again, there
has to be some source of power, since a negative resistance sources
power to whatever is connected to it.



I was thinking the neon lamp could display negative resistance at certain
voltage/current ranges. The tunnel diode can also.
Not sure right off it they are active devices or not without looking at some
old information I have.

"Roy Lewallen" wrote in message
...
I see from my records that you have EZNEC. Open EZNEC example file
4square.ez, and click Src Dat. You'll see that element 1 has a negative
resistance. In the main EZNEC window, open the File menu, then click
Edit Current Antenna Notes. This contains an explanation of the
phenomenon. It can happen only when there are multiple sources, since
the power to create the negative resistance at one source comes from
power supplied by others. If you use a feed system to distribute power
to various elements from a single source, the resistance at that source
can never be negative.

Okay. I have only the one source, so I should never see a negative
resistance is what I read from your reply.

Active devices can also exhibit a negative resitance, but again, there
has to be some source of power, since a negative resistance sources
power to whatever is connected to it.

Yes. That's why I made sure I mentioned passive. I meant no active devices.

Your result could be due to measurement error, erroneous setup or
procedure, or, if you're measuring an antenna, RF on the antenna from
outside sources. The latter makes antenna impedance measurement
difficult without a sharply tuned detector, which the 8405A doesn't have.

Error was my assumption as well. Thanks for your input, Roy. I'll take a
look at EZNEC as you suggested.

John




Roy Lewallen, W7EL

The other John Smith wrote:

Hi, Gentlemen -

I have an HP 8405A Vector Voltmeter. I have the App Note "Measuring
Complex
Impedance" by HP (thanks to Wes Stewart). I have made a few measurements
following (maybe) the HP paper. I have probably been sloppy with the
setup
and procedure because I came up with a negative resistance a couple of
times.

The HP paper discusses negative resistance but does not mention where
one
would encounter it. I assume it would normally be associated with active
device measurements. Is this true? If I measure a (passive) load and
find
that the resistance is negative, shouldn't I doubt my setup and/or
procedure? Is there any condition where a passive, real-world, network
could
show a negative resistance?

By the way, I have not gotten a negative resistance reading while using
a
dual directional coupler. I have gotten the negative resistance reading
only
when using the power splitter method. So far.

Thanks for any insight on this.

John KD5YI

"Ralph Mowery" wrote in message
ink.net...
Active devices can also exhibit a negative resitance, but again, there
has to be some source of power, since a negative resistance sources
power to whatever is connected to it.



I was thinking the neon lamp could display negative resistance at certain
voltage/current ranges. The tunnel diode can also.
Not sure right off it they are active devices or not without looking at
some
old information I have.


Hi, Ralph -

Yes, I am familiar with tunnel diodes from way back in the 60's. They were
GE devices as I recall. In fact, until about 10 years ago, I had an even
older schematic from an EDN magazine article that showed how to connect two
NPN transistors together to emulate a tunnel diode. Looked sort of like a
bistable connection, as I recall.

Anyway, that's why I limited my statements to passive devices although, as
Roy pointed out, I did not state that I had only one source.

Thanks.

John

"Richard Fry" wrote in message
...
"The other John Smith" wrote
Is there any condition where a passive, real-world, network
could show a negative resistance?
____________

"Negative towers' exist in some MW broadcast directional arrays. Here is
a
link to a paper about it: http://www.qsl.net/km5kg/negative.htm. Don't
know if that addresses your situation or not...

RF


Thanks, Richard. I'll take a look.

John

To test a device, connect a resistor to its terminals. If the resistor
gets warm, you have a negative resistance.

The neon lamp and tunnel diode exhibit a negative *dynamic* resistance.
That it, the *slope* of the V/I curve is negative at some parts of the
range. They don't exhibit a true negative resistance -- that is, under
no conditions will you find the value of V/I to be negative.

Roy Lewallen, W7EL

Ralph Mowery wrote:

I was thinking the neon lamp could display negative resistance at certain
voltage/current ranges. The tunnel diode can also.
Not sure right off it they are active devices or not without looking at some
old information I have.

We had to stop using 8405s in the field many years ago due to the external
RF problem Roy describes. Many of the newer analyzers use synchronous
detection which reduces the problem, unless the RF signal level from the
antenna exceeds the dynamic range of the instrument. Put a wideband power
meter or true RMS RF voltmeter at the measurement terminals and see how much
voltage/power is there from your antenna. You may have to wait for a
blackout to make measurements if the RF is from sky wave signals. If it is
ground wave from MW broadcast stations, move.

--
Crazy George
Remove N O and S P A M imbedded in return address

On Fri, 20 Aug 2004 13:14:45 GMT, "The other John Smith"
wrote:

The 8405 would be unable to lock onto any frequency below about 200 MHz
(IIRC) due to the setting on the coarse frequency control at the time.


This is true.

I readily accept the fact that my technique was sloppy and my results showed
it. I just hope it was not due to an outside source. I can't afford to move.

But, I am new to this. My naive reasoning may be flawed. I would appreciate
any further comments on this.

Thanks.

John

I use my 8405 extensively, have two of 'em, and I've never had any indications
of negative resistance. What are the meter indications that seem to indicate
negative resistance?

Walt, W2DU

What we encountered were dynamic range problems. The instrument operates as
HP says, but if a strong enough interfering signal appears at an input, it
overloads the first(usually) active device, causing various unwanted
products to appear in band, and sometimes desensitizing the channel. Urban
problem with many other radio services nearby, and it doesn't help to be
under the landing path of USAF aircraft with their jammers still active
either.

--
Crazy George
Remove N O and S P A M imbedded in return address

On Fri, 20 Aug 2004 14:48:41 -0500, "Crazy George"
wrote:

What we encountered were dynamic range problems. The instrument operates as
HP says, but if a strong enough interfering signal appears at an input, it
overloads the first(usually) active device, causing various unwanted
products to appear in band, and sometimes desensitizing the channel. Urban
problem with many other radio services nearby, and it doesn't help to be
under the landing path of USAF aircraft with their jammers still active
either.
Hi George,

I understand the situation you encountered, but as I understand the 'other
John's' situation, he was sufficiently far from strong RF fields such that there
would be no problem with dynamic range. My interest is what the meters showed on
his 8405 that would be recognized as negative resistance.

Walt, W2DU

"Walter Maxwell" wrote in message
...
On Fri, 20 Aug 2004 14:48:41 -0500, "Crazy George"

wrote:

What we encountered were dynamic range problems. The instrument operates
as
HP says, but if a strong enough interfering signal appears at an input,
it
overloads the first(usually) active device, causing various unwanted
products to appear in band, and sometimes desensitizing the channel.
Urban
problem with many other radio services nearby, and it doesn't help to be
under the landing path of USAF aircraft with their jammers still active
either.
Hi George,

I understand the situation you encountered, but as I understand the 'other
John's' situation, he was sufficiently far from strong RF fields such that
there
would be no problem with dynamic range. My interest is what the meters
showed on
his 8405 that would be recognized as negative resistance.

Walt, W2DU


Hi, Walt -

I'm sorry to say that I did not keep the data. Shame on me. But I will keep
the data the next time it comes up (if it does) and I will post it here. As
I said, I have not had negative resistance show up while using the dual
directional coupler, only when using the power splitter and pads. Different
math relations were used to calculate the results from the two techniques,
so it might very well be that it will never show up again.

Our country property where the measurements were made is about 10 miles east
of Sulphur Springs, Texas, and about 80 or so miles from Dallas. There are
two radio stations there, one on 1230 kHz (1 kW) and one on 95.9 MHz (6 kW).
The nearest microwave tower is about 5 miles west of me and I don't think
I'm in the path. There are no hams closer than 8 miles to me.

The more I think about this, the more convinced I become that I flubbed the
readings or the calculations. I am learning that you cannot treat 2 meter
and 70 cm signals casually. For example, when placing a short on the load
port of the coupler for calibration purposes, I can see the phase changing
after the connector has made contact but is still being screwed down.

Thanks for your comments.

John

On Sat, 21 Aug 2004 00:30:48 GMT, "The other John Smith"
wrote:


"Walter Maxwell" wrote in message
.. .
On Fri, 20 Aug 2004 14:48:41 -0500, "Crazy George"

wrote:

What we encountered were dynamic range problems. The instrument operates
as
HP says, but if a strong enough interfering signal appears at an input,
it
overloads the first(usually) active device, causing various unwanted
products to appear in band, and sometimes desensitizing the channel.
Urban
problem with many other radio services nearby, and it doesn't help to be
under the landing path of USAF aircraft with their jammers still active
either.
Hi George,

I understand the situation you encountered, but as I understand the 'other
John's' situation, he was sufficiently far from strong RF fields such that
there
would be no problem with dynamic range. My interest is what the meters
showed on
his 8405 that would be recognized as negative resistance.

Walt, W2DU


Hi, Walt -

I'm sorry to say that I did not keep the data. Shame on me. But I will keep
the data the next time it comes up (if it does) and I will post it here. As
I said, I have not had negative resistance show up while using the dual
directional coupler, only when using the power splitter and pads. Different
math relations were used to calculate the results from the two techniques,
so it might very well be that it will never show up again.

Our country property where the measurements were made is about 10 miles east
of Sulphur Springs, Texas, and about 80 or so miles from Dallas. There are
two radio stations there, one on 1230 kHz (1 kW) and one on 95.9 MHz (6 kW).
The nearest microwave tower is about 5 miles west of me and I don't think
I'm in the path. There are no hams closer than 8 miles to me.

The more I think about this, the more convinced I become that I flubbed the
readings or the calculations. I am learning that you cannot treat 2 meter
and 70 cm signals casually. For example, when placing a short on the load
port of the coupler for calibration purposes, I can see the phase changing
after the connector has made contact but is still being screwed down.

Thanks for your comments.

John

Hi John,

I believe I said earlier that the distance to any probable source of
interference to your 8405 is sufficient to exclude any interference.

Is it possible to recall what you were measuring, and what the setup was when
yo;u obtained the negative resistance indications, such that you could repeat
it just to humor me?

And concerning the phase change when screwing down the short, it's possible
there's contamination in the screw threads, either on the short or on the
connector on the coupler. On the other hand, at UHF the phase can change
slightly between having the short placed on the coupler before tightening down
vs being completely tightened down.

I use an HP 778D coupler with N connectors, but I normally use a BNC short on an
adapter when establishing a phase and magnitude reference. When using sufficient
padding between the sig gen and the coupler I find no difference in magnitude of
the reflection between the short or open reference.

I'm curious concerning what coupler and sig gen you use. My gens are all HP, the
606A, 608E, and 612.

Walt, W2DU

"Walter Maxwell" wrote in message
...
On Sat, 21 Aug 2004 00:30:48 GMT, "The other John Smith"

wrote:


"Walter Maxwell" wrote in message
.. .
On Fri, 20 Aug 2004 14:48:41 -0500, "Crazy George"

wrote:

What we encountered were dynamic range problems. The instrument
operates
as
HP says, but if a strong enough interfering signal appears at an
input,
it
overloads the first(usually) active device, causing various unwanted
products to appear in band, and sometimes desensitizing the channel.
Urban
problem with many other radio services nearby, and it doesn't help to
be
under the landing path of USAF aircraft with their jammers still
active
either.
Hi George,

I understand the situation you encountered, but as I understand the
'other
John's' situation, he was sufficiently far from strong RF fields such
that
there
would be no problem with dynamic range. My interest is what the meters
showed on
his 8405 that would be recognized as negative resistance.

Walt, W2DU


Hi, Walt -

I'm sorry to say that I did not keep the data. Shame on me. But I will
keep
the data the next time it comes up (if it does) and I will post it here.
As
I said, I have not had negative resistance show up while using the dual
directional coupler, only when using the power splitter and pads.
Different
math relations were used to calculate the results from the two
techniques,
so it might very well be that it will never show up again.

Our country property where the measurements were made is about 10 miles
east
of Sulphur Springs, Texas, and about 80 or so miles from Dallas. There
are
two radio stations there, one on 1230 kHz (1 kW) and one on 95.9 MHz (6
kW).
The nearest microwave tower is about 5 miles west of me and I don't think
I'm in the path. There are no hams closer than 8 miles to me.

The more I think about this, the more convinced I become that I flubbed
the
readings or the calculations. I am learning that you cannot treat 2 meter
and 70 cm signals casually. For example, when placing a short on the load
port of the coupler for calibration purposes, I can see the phase
changing
after the connector has made contact but is still being screwed down.

Thanks for your comments.

John

Hi John,

I believe I said earlier that the distance to any probable source of
interference to your 8405 is sufficient to exclude any interference.

Is it possible to recall what you were measuring, and what the setup was
when
yo;u obtained the negative resistance indications, such that you could
repeat
it just to humor me?


No, I'm afraid not. I made so many measurements that weekend that it sort of
turned my brain to mush. I will set up the apparatus as I did before and
repeat some experiments. I like to be humored, too.


And concerning the phase change when screwing down the short, it's
possible
there's contamination in the screw threads, either on the short or on the
connector on the coupler. On the other hand, at UHF the phase can change
slightly between having the short placed on the coupler before tightening
down
vs being completely tightened down.


The phase change is not a problem, just an observation. After thinking about
it, it makes perfect sense. It was due to the shortening of the path as the
shorted connector was screwed on. The change was not large, maybe a degree
or two. It just surprised me that the setup was that sensitive.


I use an HP 778D coupler with N connectors, but I normally use a BNC short
on an
adapter when establishing a phase and magnitude reference. When using
sufficient
padding between the sig gen and the coupler I find no difference in magnit
ude of
the reflection between the short or open reference.


I use a 11570A kit which includes a power splitter, two probe Ts, two
terminators and a short. I also add two attenuators per the HP application
note. All are type N connectors.


I'm curious concerning what coupler and sig gen you use. My gens are all
HP, the
606A, 608E, and 612.


Recently I have replaced the splitter and attenuators with a Narda dual
directional coupler. It is with this setup that I seem to be moderately
successful. My generator is an HP 3200B VHF Oscillator.

John

"Walter Maxwell" wrote in message
...
On Sat, 21 Aug 2004 00:30:48 GMT, "The other John Smith"

wrote:

Is it possible to recall what you were measuring, and what the setup was
when
yo;u obtained the negative resistance indications, such that you could
repeat
it just to humor me?


Okay, I just found a tiny note on a piece of paper which says:

"B1/A1 = .90 at 180 degrees reference (shorted)"

and

"16 and 1/2 inches of RG58A (shorted) gives B2=0.74, A2=0.81, at -110
degrees"


This works out to -0.56 - 35i. Unfortunately, my note does not indicate the
nature of the load.

It looks like the real part goes negative if A1B2/A2B1 1. Yes, I think I
see it now. If the reflection coefficient is greater than one, that
indicates more is being reflected than is being supplied -- meaning I have a
source at the supposedly shorted end of the coax. Ah ha! I am now very
confident that I either misread the instruments or misadjusted something.

Anyway, I think I have the answer I was seeking. Except under extraordinary
circumstances, I should never get a negative real part answer.

My thanks to the contributors of this thread. It sure helps to have others
to talk things over with.

John

John, negative resistance is indicated when the magnitude of the reflection
coefficient is greater than one.
The solution of the impedance equation for rho 1 yields negative values of
resistance. This lends a lot
of insight into possible causes:

The calibration for open or short is slightly lossy. If a measured load
has a little more reflection than the
calibration loads, it would indicate as negative resistance.

Numerical errors creep into the results. The Smith chart is extremely
sensitive around the periphery.
A few tenths of a dB. result in large movements at the edge of the chart.
Thus, small calibration or
computational errors can result in crossing over the rho=1 circle.

And, as mentioned, detection of external RF fields increasing the output
of the reflection detector
past the calibration value can result in a measured value for rho 1.

-- Tom, N5EG






"The other John Smith" wrote in message
ink.net...

The HP paper discusses negative resistance but does not mention where one
would encounter it. I assume it would normally be associated with active
device measurements. Is this true? If I measure a (passive) load and find
that the resistance is negative, shouldn't I doubt my setup and/or
procedure? Is there any condition where a passive, real-world, network
could
show a negative resistance?



John KD5YI

Walter Maxwell wrote:

On Sat, 21 Aug 2004 16:31:10 -0500, "The Other John Smith"
wrote:


Well, John, there's nothing wrong with your head, except perhaps you're not
really acquainted with the characteristics of shorted RG58's. RG58's are notable
for their storage of heat. After they are energized, and the souce disconnected,
then as the coax cools down and the heat is radiated, a new emf is generated
causing current to flow in the reverse direction. The coax is now a new source
and the power you measured with the 8405 is indicated as negative resistance.

Walt

PS--I once thought I'd become a standup comedian, but the competition was too
great.


Was the RG58 vertical at the time? If you have the shorted up when
vertical, the charge gets pulled back down, causing partial, or during
times of gravitational anamolies, more than complete, cancelation of the
forward looking, i.e "liberal", current, which we all know causes Pointy
Vector to swerve to the right. If you hang the shorted end down, the
reverse occurs, causing more current, or "conservative" current, and
swings the Pointy Vector to the left.

One controversial way to deal with this, is to spin the shorted section
in the horizontal plane, and then we seem to get charge flowing with no
real direction calling itself - "Nader".

tom
K0TAR

Tom Ring wrote:

snip
Vector to swerve to the right. If you hang the shorted end down, the
reverse occurs, causing more current, or "conservative" current, and
swings the Pointy Vector to the left.


Sorry, should have read "causing more current, or "conservative" current
to be canceled".

tom
K0TAR

On Sat, 21 Aug 2004 20:46:15 -0500, Tom Ring wrote:

Tom Ring wrote:

snip
Vector to swerve to the right. If you hang the shorted end down, the
reverse occurs, causing more current, or "conservative" current, and
swings the Pointy Vector to the left.


Sorry, should have read "causing more current, or "conservative" current
to be canceled".

tom
K0TAR

Verily,

Walt, W2DU

On Sat, 21 Aug 2004 18:07:16 GMT, "The other John Smith"
wrote:

I tried sending you an email on this but you've spoofed your address
so bad I guess you didn't get it.

Pity.

"Walter Maxwell" wrote in message
...
On Sat, 21 Aug 2004 20:46:15 -0500, Tom Ring wrote:

Tom Ring wrote:

snip
Vector to swerve to the right. If you hang the shorted end down, the
reverse occurs, causing more current, or "conservative" current, and
swings the Pointy Vector to the left.


Sorry, should have read "causing more current, or "conservative" current
to be canceled".

tom
K0TAR

Verily,

Walt, W2DU


Snicker! You guys are a hoot!

John, KD5YI

"TOM" wrote in message
...
John, negative resistance is indicated when the magnitude of the
reflection
coefficient is greater than one.
The solution of the impedance equation for rho 1 yields negative values
of
resistance. This lends a lot
of insight into possible causes:

The calibration for open or short is slightly lossy. If a measured
load
has a little more reflection than the
calibration loads, it would indicate as negative resistance.

Numerical errors creep into the results. The Smith chart is extremely
sensitive around the periphery.
A few tenths of a dB. result in large movements at the edge of the chart.
Thus, small calibration or
computational errors can result in crossing over the rho=1 circle.

And, as mentioned, detection of external RF fields increasing the
output
of the reflection detector
past the calibration value can result in a measured value for rho 1.

-- Tom, N5EG


Yes, I see that now, Tom. Thanks.

John, KD5YI

"John Smith" kd5yiatmindspringdotcom wrote in message
...

"TOM" wrote in message
...
John, negative resistance is indicated when the magnitude of the
reflection
coefficient is greater than one.
The solution of the impedance equation for rho 1 yields negative
values
of
resistance. This lends a lot
of insight into possible causes:

The calibration for open or short is slightly lossy. If a measured
load
has a little more reflection than the
calibration loads, it would indicate as negative resistance.

Numerical errors creep into the results. The Smith chart is
extremely
sensitive around the periphery.
A few tenths of a dB. result in large movements at the edge of the
chart.
Thus, small calibration or
computational errors can result in crossing over the rho=1 circle.

And, as mentioned, detection of external RF fields increasing the
output
of the reflection detector
past the calibration value can result in a measured value for rho 1.

-- Tom, N5EG


Yes, I see that now, Tom. Thanks.

John, KD5YI

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

Measured values of Rho greater than 1.0 are not necessarily due to
measurement errors. They can be true. The true theoretical maximum value of
Rho is 1 + Sqrt(2) which occurs on transmission lines only at extremely low
frequencies and with a purely inductive terminating impedance. It arises due
to resonance between line impedance and the termination and cannot be of any
consequence at RF.

Values of Rho greater than 1.0 are mathematically legitimate as may be
demonstrated by exact computer programs when presented with exact data. In
any case, the Smith Chart is inherently inexact at the lower frequencies in
addition to the user's difficulties of reading from it.

The reflection coefficient can quite easily lie in any of the four 90-degree
quadrants. If the real component of the vector is negative it may be thought
to correspond somewhere to an imaginary negative resistance. But when Rho is
correctly used in line input and output impedance calculations it always
gives the correct answers. ie., negative resistances cannot exist.
Resistances are always lossy.
----
Reg, G4FGQ

"Wes Stewart" wrote in message
...
On Sat, 21 Aug 2004 18:07:16 GMT, "The other John Smith"
wrote:

I tried sending you an email on this but you've spoofed your address
so bad I guess you didn't get it.

Pity.

Hi, Wes. Yes it is a pity that we have to go to so much inconvenience to
keep spam to a tolerable level. Sorry about that.

I tried to email you, too, but you may have not checked it by the time I
post this. You can reach me through the ARRL reflector as kd5yi at arrl dot
net.

Thanks.

John, KD5YI