Today I was making some measurements with my MFJ-259B. From the manual,
I read:

"The MFJ-259B measures reactance, and converts reactance to
capacitance. The MFJ-259B
can not determine if the reactance is actually inductive or capacitive.
You can usually determine
the type of reactance by adjusting frequency. If frequency is increased
and reactance (X on the
display or Impedance on the meter) decreases, the load is capacitive at
the measurement
frequency. If frequency is reduced and reactance decreases, the load is
inductive at the
measurement frequency."

Then I looked over the measurements that I plotted out. If I look at
the graph of the reactance, the amplitude of the reactance will
increase and then begin to decrease as frequency increases. If I try to
follow the tip given in the MFJ manual, the reactance can appear as
both capactive and reactive before it ever crosses zero. For instance
if I follow the graph it rises and then starts to fall - according to
the manual this would mean the upside is inductive and the downside is
capacitive. But what I don't understand is how that is possible - it
seems to me it would have to cross the integer value zero (0) to change
from capacitive reactance to inductive reactance?

What concept am I screwing up here?

-Scott, WU2X

Here are the values I measured:


Freq R X SWR
--------------------------------------
144 44 2 1.1
144.25 46 4 1.1
144.5 55 7 1.1
144.75 60 7 1.2
145 68 0 1.3
145.25 78 0 1.4
145.5 79 0 1.4
145.75 73 11 1.5
146 63 21 1.5
146.25 53 23 1.5
146.5 46 21 1.5
146.75 40 18 1.5
147 37 10 1.5
147.25 35 8 1.4
147.5 37 6 1.3
147.75 41 10 1.3
148 48 15 1.3
148.25 59 19 1.4
148.5 75 16 1.6
148.75 90 0 1.7
149 96 6 1.9

On 19 Dec 2006 22:41:37 -0800, wrote:

Here are the values I measured:

Hi Scott,

This looks like one sick puppy. Is it a dipole or a monopole? Try
adding some feedline decoupling/choking (and again a quarter wave away
from the feed point) and repeating your measurements.

73's
Richard Clark, KB7QHC

wrote:

What concept am I screwing up here?

-Scott, WU2X

Here are the values I measured:


Freq R X SWR
--------------------------------------
144 44 2 1.1
144.25 46 4 1.1
144.5 55 7 1.1
144.75 60 7 1.2
145 68 0 1.3
145.25 78 0 1.4
145.5 79 0 1.4
145.75 73 11 1.5
146 63 21 1.5
146.25 53 23 1.5
146.5 46 21 1.5
146.75 40 18 1.5
147 37 10 1.5
147.25 35 8 1.4
147.5 37 6 1.3
147.75 41 10 1.3
148 48 15 1.3
148.25 59 19 1.4
148.5 75 16 1.6
148.75 90 0 1.7
149 96 6 1.9

Hi Scott, It probably means you are measuring the impedance through a
length of lossy coax. If you do the sweep using an electrical 1/2
wavelength of coax at say 146 MHZ you will probably see more of what
you would expect. I think the MFJ manual gives instructions for cutting
a 1/2 wave xmission line. If you have EZNEC you could do an SWR sweep
of your particular antenna and see what the impedances should be. Then
take the MFJ readings right at the antenna or through a multiple of 1/2
wave coax. The results should pretty closely agree.

If you still have problems, then there may be some strong RF present
that is disturbing the measurments, or YOUR presense in the near field
of the antenna could be upsetting things. If you are trying to measure
a voltage fed antenna like a J pole, if you are physically close to the
antenna it will foul things up. Good Luck.

Gary N4AST

Looks like coax to me too....

Another problem is that the MFJ-259 isn't a very good VHF impedance
meter. The resistance and reactance readings aren't going to be at all
reliable on 2m, unless the impedance is close to 50 ohms, resistive.

I recently recalibrated mine since it was giving weird results on
purely reactive loads. The calibration procedure checks resistive
loads of 12.5, 50, 75, 100, and 200 ohms. I used SMT resistors on BNC
connectors as decent loads.

After I did the calibration, the 200 ohm load at 144MHz reads something
like R=80 X=50!

In the VHF range, even a good N-connectorized 50 ohm termination shows
a little tiny bit of reactance. R=49, X=7 or something like that.

From 1.7-30 MHz my 200 ohm load reads R=200, X=0. Up to about 60 MHz,
a little "reactance" creeps in, some of which may be genuine
(connectors and adapters), but much of which is simply due to the
bridge being substantially off up there.

If you want to do experiments with the '259's capabilities and plot
resistance and reactance curves, work on HF instead...

That said, if you're doing this through coax, try it again without and
your results might make more sense... you are "close" to 50 ohms
resistive...

Dan

I thought the SWR curve looked strange as well. The SWR has a weird
curve across the entire band. The antenna is actually a 4 element quad.
Its fed with a W2DU style VHF 1:1 balun. The coax is a short run of
RG-8X. The RG-8X uses PL-259s on both ends and I soldered a SO-239 onto
the coax for the balun (see pictures). Any issue with using PL259s and
a SO239?

Here is a picture of the feed system:
http://www.wu2x.com:8888/usenet/P1010062.JPG

Here is a picture of the antenna under test:
http://www.wu2x.com:8888/usenet/P1010058.JPG

This is the data graphed out, so its easier to see:
http://www.wu2x.com/usenet/ReadingA-ReadingB.htm

I took measurements with it fed for vertical polarization and
horizontal polarization. The results are both shown on the graphs.

-Scott, WU2X


Richard Clark wrote:
On 19 Dec 2006 22:41:37 -0800, wrote:

Here are the values I measured:

Hi Scott,

This looks like one sick puppy. Is it a dipole or a monopole? Try
adding some feedline decoupling/choking (and again a quarter wave away
from the feed point) and repeating your measurements.

73's
Richard Clark, KB7QHC

Your coax is what, 10-12 feet long? That's pushing a couple of
wavelengths on 2m, but isn't quite there.

To read the impedance of the antenna directly and get a good set of
resistance and reactance curves clearly, you need probably less than a
foot of coax, or an exact multiple of an electrical half wavelength (so
exactly 2 wavelengths * velocity factor would do it). You can use your
MFJ-259 to cut the line.

It's possible to transform your impedance readings into the antenna's
impedance if you know the (random) electrical length and loss of the
coax and are handy with a Smith chart or have some other calculating
ability.

Exact multiple of a half wave is easier... it's an impedance repeater,
aside from the loss in the line. To that end, I'd use an exact
multiple of half wave of better coax... RG-8 or 9913 or something.

73,
Dan

On 20 Dec 2006 09:16:47 -0800, wrote:

I thought the SWR curve looked strange as well. The SWR has a weird
curve across the entire band. The antenna is actually a 4 element quad.
Its fed with a W2DU style VHF 1:1 balun.

Hi Scott,

Well, I can plainly see the BalUn in the first picture. To be
completely decoupled a second one a quarterwave away on the feedline
would help, but the numbers look far more whacked out than that might
cure (because it probably isn't the problem to cure).

The coax is a short run of
RG-8X. The RG-8X uses PL-259s on both ends and I soldered a SO-239 onto
the coax for the balun (see pictures). Any issue with using PL259s and
a SO239?

I notice from the picture you don't seal anything. Those connectors
are leakier than a sieve.

Aside from these comments, perhaps you should validate your meter by
measuring known loads. It sounds simple, but in the VHF it is not.
Common garden variety components like resistors and capacitors have
values that diverge greatly from their markings. So, some care in
their selection (or construction) is warranted.

Anyway, grabbing components out of your junk-box would serve as a
first pass attempt, but it shouldn't end there.

73's
Richard Clark, KB7QHC

Its not that I am trying to get the exact feedpoint impedance values
from the antenna. I understand how 50 Ohm coax affects the values that
I've measured. I was originally trying understand how to determine if a
reactance
value was capacitive or inductive.

I do see now that when I use NEC2 to plot the values for this antenna,
the resistance and reactance appear more linear across the frequency
range I measured. I see if I apply the logic of the MFJ manual to the
plot from NEC2, it makes sense and that my data is all over the place.
For what reasons I do not know. I do not think the feedline is
radiating. Maybe it is just the meter. Or perhaps some larger deviation
from 50 ohms in the coax and connectors.

And Richard - the antenna doesn't stay outside, that is why its not
sealed. This is just rigged up for testing.

-Scott, WU2X

wrote:
Your coax is what, 10-12 feet long? That's pushing a couple of
wavelengths on 2m, but isn't quite there.

To read the impedance of the antenna directly and get a good set of
resistance and reactance curves clearly, you need probably less than a
foot of coax, or an exact multiple of an electrical half wavelength (so
exactly 2 wavelengths * velocity factor would do it). You can use your
MFJ-259 to cut the line.

It's possible to transform your impedance readings into the antenna's
impedance if you know the (random) electrical length and loss of the
coax and are handy with a Smith chart or have some other calculating
ability.

Exact multiple of a half wave is easier... it's an impedance repeater,
aside from the loss in the line. To that end, I'd use an exact
multiple of half wave of better coax... RG-8 or 9913 or something.

73,
Dan

wrote in message
oups.com...
Looks like coax to me too....

Another problem is that the MFJ-259 isn't a very good VHF impedance
meter. The resistance and reactance readings aren't going to be at all
reliable on 2m, unless the impedance is close to 50 ohms, resistive.

I recently recalibrated mine since it was giving weird results on
purely reactive loads. The calibration procedure checks resistive
loads of 12.5, 50, 75, 100, and 200 ohms. I used SMT resistors on BNC
connectors as decent loads.

After I did the calibration, the 200 ohm load at 144MHz reads something
like R=80 X=50!

In the VHF range, even a good N-connectorized 50 ohm termination shows
a little tiny bit of reactance. R=49, X=7 or something like that.

From 1.7-30 MHz my 200 ohm load reads R=200, X=0. Up to about 60 MHz,
a little "reactance" creeps in, some of which may be genuine
(connectors and adapters), but much of which is simply due to the
bridge being substantially off up there.

If you want to do experiments with the '259's capabilities and plot
resistance and reactance curves, work on HF instead...

That said, if you're doing this through coax, try it again without and
your results might make more sense... you are "close" to 50 ohms
resistive...

Dan

I tried using one of those MFJ analyzers on 2M and it didnt work so great.
One of the big problems wa that my presence would effect the reading. I
could see the reading changed as I walked toward the antenna. I had the same
problem with cable length. An SWR meter became more valuable.

wrote in message
oups.com...
Its not that I am trying to get the exact feedpoint impedance values
from the antenna. I understand how 50 Ohm coax affects the values that
I've measured. I was originally trying understand how to determine if a
reactance
value was capacitive or inductive.

I do see now that when I use NEC2 to plot the values for this antenna,
the resistance and reactance appear more linear across the frequency
range I measured. I see if I apply the logic of the MFJ manual to the
plot from NEC2, it makes sense and that my data is all over the place.
For what reasons I do not know. I do not think the feedline is
radiating. Maybe it is just the meter. Or perhaps some larger deviation
from 50 ohms in the coax and connectors.

And Richard - the antenna doesn't stay outside, that is why its not
sealed. This is just rigged up for testing.

-Scott, WU2X

wrote:
Your coax is what, 10-12 feet long? That's pushing a couple of
wavelengths on 2m, but isn't quite there.

To read the impedance of the antenna directly and get a good set of
resistance and reactance curves clearly, you need probably less than a
foot of coax, or an exact multiple of an electrical half wavelength (so
exactly 2 wavelengths * velocity factor would do it). You can use your
MFJ-259 to cut the line.

It's possible to transform your impedance readings into the antenna's
impedance if you know the (random) electrical length and loss of the
coax and are handy with a Smith chart or have some other calculating
ability.

Exact multiple of a half wave is easier... it's an impedance repeater,
aside from the loss in the line. To that end, I'd use an exact
multiple of half wave of better coax... RG-8 or 9913 or something.

73,
Dan


It may be YOU that is causing the vaules to be so erratic. Your pressence
can detune the antenna.
10- 12 feet is too close.