Feature request - Measuring bal. loads w/o balun

[dl4raj]

Hi Bob,

what do you think of adding a unique feature to AIM:
the capability of measuring antennas with balanced
feeders,i.e. generally speaking, balanced loads over frequency
*without* a balun?
We all know the limits of baluns when it comes to
higher impedances,particularly if they contain considerable
reactance.
The technique which is straightforward is described in W9CF's article
"Putting a Balun and Tuner Together" on the last few pages.

fermi.la.asu.edu/w9cf/articles/balun.pdf

This method has already been described in the manual of
the General Radio RF Bridge 916-A which was built
in the 50ies!

One would have to do just three scans, the AIM
software would have to do the quite simple math
(but for every scan point) and then create a trace
for the true differential impedance.
Until now I use to to this with a programmable HP handheld
calculator for single frequencies when I measure tuned
doublet/open feeder systems,a tedious procedure.
It would be really great to have this done by AIM
over the whole frequency range of interest.
Moreover one could see how unbalance changes
over frequency if all three scans were presented in a
overlay mode.
Comments welcome!

73
Clemens
DL4RAJ

[k6mhe]

According to the folks at Agilent you cannot measure a balanced load with an unbalanced instrument.

Here’s a partial quote form Agilent Application Note AN 346-2:
“….. A balanced circuit cannot be directly measured with an unbalanced measurement instrument because of the difference in their configuration.
When measuring balanced circuits, the unbalanced measuring instrument requires a balun (balanced to unbalanced) transformer….” cp.literature.agilent.com/litweb/pdf/5091-4480E.pdf )

That said, I have had fairly good success measuring balanced lines at the lower HF frequencies by using a common mode choke/balun (similar to MFJ’s analyzer balun but mine has a slightly higher common mode impedance) and floating the AIM above ground using a battery for power and a bluetooth wireless RS232 connection.

As I recall, the technique for measuring the amount of unbalance in W9CF's article involved measuring the parallel lines one line at a time referenced to ground. Because that is not measuring a balanced load it therefore would not require a balun, however, I feel it still would be advisable to have the AIM “floating”.

Perhaps Larry, W0QE, might jump in here with some guidance?

73,
Danny, K6MHE

[dl4raj]

Danny,

>“….. A balanced circuit cannot be directly measured with an >unbalanced measurement instrument because of the difference in >their configuration.

I believe the keyword is is *directly*.
The technique described in W9CF's article derives from simple
circuit theory and as I've mentioned, was well known already around 1950 and described in the manuals of the GR impedance bridges.
Indeed it is an *indirect* way.

73
Clemens
DL4RAJ

[w0qe]

Hi Clemens and Danny,

I am in no way an expert on this topic but I have played around making some of these measurements several times. Let me make a few comments which may help or maybe they will just create more questions.

1.) The common mode impedance is not really important. What is important is the common mode current that flows which is not only due to the common mode impedance of the feedline/antenna alone but also the common mode impedance of the transmitter/tuner that connects to it. For example a true balanced antenna tuner has a high common mode impedance and the common mode impedance of the lin is not very important.

2.) If you do not have "real ground" which to make the measurements against but say some impedance above ground then the numbers you measure will be in error.

3.) There are 2 fundamental uses for common mode chokes/baluns. The first is on the transmission line that is being excited directly and which large common mode currents may flow. A balun used for this purpose needs to be designed which very low loss material such as #61, powdered iron, or even air and needs the common mode impedance to be either nearly all inductive or capacitive so it doesn't heat up. The second use is to stop common mode current on wires that are coupled to from the transmitting source. This use has much lower common mode currents (even though the stereo, appliance, etc. is affected) and can use common mode chokes wound on #73, #31, or #42 materials and much of the common mode impedance is resistive. Here low Q and big R are our friends.

4.) The current that flows, pre common mode choke, is a function of the common mode driving impedance of the wire and the common mode terminating impedance of the equipment. Putting in a common mode choke adds a 3rd impedance which is what is counted upon to reduce the common mode current. This is why sometimes a single small core with one turn will solve a problem and other times why many turns on a couple of cores are needed. Also always know the magnitude of the "enemy". If running 10 watts causes an appliance to be interfered with and if you want to run 1000 watts then you will need to reduce the common mode current by 20dB or a factor of 10 which may need choking impedances in the range of several thousand ohms. However if interference shows up at 500 watts then you only need to reduce the common mode current by 3dB or a factor of 1.4 which may be accomplished by less than 50 ohms of choking impedance.

5.) In my opinion since several things affect the common mode current the easiest thing to do is to just measure it directly with an EMI type of current transformer and some kind of RF power meter. Then you can put in a one turn common mode choke (which was previously measured) and see how much the current was reduced and then know the sum of the common mode impedances of both the wire and the appliance/transmitter.

Does this make sense? As far as the feature request for the 4170 is concerned I kind of question its value but I may be wrong.

Comments?

73, Larry, W0QE

[chuckshinn]

Clemens; I have wanted to use my AIM for similar tests. In my case I am driving a balanced doublet with 500 ohm homemade balanced feed line. I wanted to check balance in the system over 3 to 30 MHz. To do this I first swept my antenna with hot lead on (arbitrarily) the left lead of the balanced antenna feeders. Then the right. Comparing the visual results proved unsatisfactory unless done for a small frequency band but a thought occurred to me. Grab the resultant csv files and import into XL. Do the math there and plot in XL the difference plot would be an easy plot to make once the math is done. I agree that a method for measuring balanced loads would be pretty handy but I am wondering how many would use it?

To do this measure I had previously isolated, as best I can, the ins and outs of the AIM unit. Fully choked input line as well as floating the backside support lines. Doing the test this way does not eliminate the chokes or baluns but it takes some tedium out of making a comparison.

I don't know how this method would work if done internally in AIM software with no isolation of cables/AIM box. I am assuming that is the root of your ask here.

Best, Chas

[dl4raj]

Larry,

your comments were interesting but I'm afraid to say missed the point,except paragraph 2)

Apr 23, 2009 20:25:58 GMT -5 w0qe said:

Hi Clemens and Danny,

.........
2.) If you do not have "real ground" which to make the measurements against but say some impedance above ground then the numbers you measure will be in error.
..........................

73, Larry, W0QE

If you refer to the same point which you define as "ground"
for all three measurements (see W9CF's article) the error
of non-ideal rf ground is cancelled out.

The point was how to measure the true differential mode
impedance of a balanced load like a dipole fed by a balanced line
with an unbalanced device like the AIM.

73
Clemens
DL4RAJ

[dl4raj]

Hi Chas,

>Grab the resultant csv files and import into XL.
> Do the math there and plot in XL the difference
> plot would be an easy plot to make once
> the math is done.

There is an excel based program by Dan Maguire
which works together with AIM although unfortunately
it's no longer supported by him.
It creates all files in excel format.
Now one would have to know
(or to know someone who knows) how to do the math
in a kind of batch procedure for every scan point in excel
and then create a trace out of the results.

>I agree that a method for measuring balanced loads
>would be pretty handy but I am wondering how
> many would use it?

Maybe a lot of those who buy a self contained battery
operated earth free antenna analyzer for just this reason.
I believe that a number of Aim users underestimate the
error introduced by the fact that Aim is not earth free.
Chokes are of limited use.
Imagine a balanced load of say 10 -j2000Ohm.
A sufficient choke should have a choking impedance
of *at least* 8k to 10k over the whole sw range,
a virtually impossible task.

The bluetooth solution seems costly and bulky to me.

73
Clemens
DL4RAJ

[w0qe]

Ok, I guess I missed the point. I went back a re-read Kevin's article. I tried making these measurements on an antenna at home with my 4170 and my HP VNA. The resulting measurements seemed pretty close but after putting them into the formulas the answers were dissapointingly different. I made the 4170 measurements again the next day and came up with different results again. I'm not sure what is going on but I think it is an accuracy issue.

You are correct (I was wrong) that mathematically you can use any reference point for the measurement but look at the sensitivities of the numbers you get. The Zu = 4 * Zc^2 * Za * Zb * (Za - Zb) / S term in the W9CF article has 4 numbers multiplied together and then multiplied by the (Za - Zb) which needs to be measured very accurately. Just assume that all the Z terms were about 100 ohms then the numerator in the equation would be 100*100*100*100*(100-100) = 0 but what if you measured Za and Zb with a 0.1% error (which is miniscule). Now Zu = 100*100*99.9*100.1*(100.1-99.9) = 20 million. I will try to remake the measurements again sometime.

As to my other comments missing the point I still feel that the balance of the antenna is not what is important. What is important is what common mode current flows relative to the differential current.

73, Larry, W0QE

[dl4raj]

Apr 27, 2009 14:29:45 GMT -5 w0qe said:

Just assume that all the Z terms were about 100 ohms then the numerator in the equation would be 100*100*100*100*(100-100) = 0 but what if you measured Za and Zb with a 0.1% error (which is miniscule). Now Zu = 100*100*99.9*100.1*(100.1-99.9) = 20 million. I will try to remake the measurements again sometime.

Probably you've not done the complete calculation.

I've calculated an example (with some unbalance)with:
-----100-------.--------150------.
|
50
|
ground

(the 50 Ohm resistor to ground should be connected to the
middle between the 100R and 150R but for some reason
I can't manage to get it placed there in the picture!)

You get
Za =137,5Ohm
Zb =183,3 Ohm
Zc = 110 Ohm
Zd = 250 Ohm
Zu = 50 Ohm
Zd and Zu are calculated by the formulas presented in W9CF's
article and the result can be easily verified in this simple example
(Zd =100R + 150R =250R; Zu =150R-100R = 50R).
Then I've made variations of + and - 10% for Za,Zb,Zc.
You get 8 different combinations.
The change in Zd was:
-3,8%,+5,2%,-6,1%,+6,5%,-10%,+10%,
-18% (=the second worst case for Za-10%,Zb-10%,Zc+10%)
and
+25%(=worst case for Za+10%,Zb+10%,Zc-10%)
The two worst cases are quite unlikely to occur.
So it's safe to say that the calculated value for Zd is as at least as accurate as the measurements of Za,Zb,Zc.

>As to my other comments missing the point I still feel that the
> balance of the antenna is not what is important.
> What is important is what common mode current
> flows relative to the differential current.

From an operational point of view I fully agree with you.
However,for an optimum design of a tuning network
for a balanced fed antenna it is very helpful
to know the true differential mode impedance
(I'm not talking of the antenna's balance).
The measurement of the antenna's differential mode
impedance is falsified by a non-floating antenna analyzer.

73
Clemens
DL4RAJ

---