Balun measurements

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by VK2FXXX, Feb 13, 2011.

Thread Status:
Not open for further replies.
ad: L-HROutlet
ad: l-rl
ad: Subscribe
ad: Left-3
ad: L-MFJ
ad: abrind-2
ad: Left-2
  1. G3TXQ

    G3TXQ Ham Member QRZ Page

    The safe option is to design for the maximum achievable Resistive CM impedance across the operating frequency range.

    Steve G3TXQ
  2. VK1OD

    VK1OD Ham Member QRZ Page

    Rick, let us consider two condigurations:

    1. tx : balun : twowireline : dipole; and
    2. tx : coax : balun : dipole.

    If both configurations are identical in every detail other than the differences above, if you analyse the common mode current path, it is the same (apart from the balun location). Sure, the common mode 'conductor' formed of the openwire is a little different to the outside of the coax, but not sufficient to make much difference to the result.

    Why then is there an obvious place to put the balun, why wouldn't the balun in 2 have similar effect to that in 1 if it is placed at the tx end of the line. Will they be exactly the same in every scenario? Probably not, so one must be better at least some of the time, yet the placement in 1 and 2 are often assumed optimal for each configuration.

    For whatever reason, there are many Rules of Thumb that are often cited as a minimum requirement for an effective antenna system balun, here are just two:

    1. choking inductance greater than four times differential Zo; and
    2. choking impedance greater than 10 times differential Zo.
    They seem to satisfy the "pragmatitists", but are they meaningful measures?

    I often refer to 2 as Rule 500 as usage often assumes that differential Zo is 50, and therefore minimum choking impedance is 500Ω.

    More critical thinking suggests that desirable choking impedance is not tied to differential Zo in the general case, and that desirable choking impedance for many applications should be 2kΩ or more. (This of course excludes voltage baluns as their choking impedance (common mode impedance) is typically very low.)

    It can be argued that for many applications, choking impedance is a most important criteria, yet few designs or commercial products publish this data.

    Yes, a simple test that allowed comparison of baluns would be good, but a bench test of insertion VSWR of a balun (obviously using its nominal load) denies the common use of baluns on non-nominal loads in concert with an ATU where insertion VSWR is unimportant (because any impedance transformation can be adjusted out with the ATU).

    I think the practical approach is to publish common mode impedance data, R and X, over the frequency range for which the design or product is intended. I know of one commercial supplier who publishes |Z| vs f, I cannot recall commercial suppliers publishing R and X vs f.

    I am not suggesting that a balun must have high common mode impedance to be effective, just that for many applications it is a most important requirement and when that is the case, knowing device common mode impedance informs selection.

  3. K1BQT

    K1BQT Ham Member QRZ Page

    Balun Placement

    Owen, your point is well-taken with respect to balanced line. I rarely use it, so I guess I should amend my comment about balun placement to apply strictly to unbalanced coax feed. Also, I've been doing quite a bit with OCFDs lately, so I guess I follow the maxim, "You can never be too rich or have too much common-mode suppression".:)
  4. VK1OD

    VK1OD Ham Member QRZ Page


    That case is often raised as a matched line case. Of course, to make a matched termination of the single wire against ground of the common mode conductor, the termination needs to be at the end of the line, ie connected to the ground. The issue is making a practical connection to the 'real' RF ground.

    In any event, terminating the line does not eliminate the common mode current, is merely eliminates the standing wave.

    Your example makes assumptions about the relative magnitude of the common mode current component Icm and the differential componenets Id1 and Id2, Icm being about 50% of the others.

    In the general case, the common mode standing wave pattern and differential standing wave patterns are not necessarily aligned, in fact due to:

    1. the common mode standing wave pattern being strongly influenced by the location and impedance of the common mode ground connection;
    2. the differential mode standing wave pattern being determined by the load end differential impedance; and
    3. likely different velocity factors for each mode.
    The result is that due to the two standing waves, at some points on the line, |Icm| may be very small compared to |Id|, whilst a little further along the same line, the situation may be reversed. Very different meter readings may result from the same scenario.

    It is quite possible that |Icm| >> |Id|, and in that case, the meter needles would be in the 'MFJ balance bar'.

    I remain convinced that the measurement technique of the MFJ-835 is seriously flawed and that it does not provide a reliable indication of balance.

    The way to measure the magnitude of common mode current is to algebraically sum the line currents, as is done in the MFJ-854 if both conductors are passed through the current transformer.

  5. VK2FXXX

    VK2FXXX Ham Member QRZ Page

    Gday again
    Owen says
    These are the reasons I am trying to find a way for the ham without a $$$ VNA to be able to determine the effectiveness of their baluns.
    Even measurements with a VNA dont guarantee the thing will be effective,Owen I take your point about peoples efforts to assess baluns and how difficult it is,but what is a ham to do?
    For example I made a balun a while ago ,following a Drew Diamond article. It was a ferrite rod type.
    AS I was winding the thing ,I thought to myself "self,how do I know this contraption is going to do the job intended" I figured we just trust a well respected amateurs judgement,There was no impedance or R and X data .I am fortunate In that I just happened to have, the exact same toroid that you have measured ,the LO1238. I can make the balun to the same specs as your article and be reasonably confident of the R,X and Z.
    However when I stick it at the feed-point of my coax fed dipole, how do I know Its not just a decoration.
    This is the reasoning behind the experiment I proposed to do ,but as I now understand the test fixture I proposed( horizontal coax with cm and meter) will not represent my operational antenna ,so all bets are off.
    So we are left with the option of measuring the CM on our operational antenna feedline at several points ,preferably along the whole feedline.
    I agree published R,X and Z data would be a great help ,but I feel It would only tell us Half the story.
    Will my experiment tell me anything about the choking impedance effect of the Balun under test?
    Isnt the meter Tom W8JI made for this purpose adequate?
    Its more in my price range!

    Im enjoying this thread
    Last edited: Feb 16, 2011
  6. VK1OD

    VK1OD Ham Member QRZ Page


    Measuring impedances at RF is not trivial because your test fixture can influence the results significantly.

    Here is a sidetrack experiment that you can try.

    Make up a little test fixture that takes the signal from a signal generator or low power transmitter, and feed it to a series resistor (say 100Ω), and another 'unknown' resistor to ground. Arrange yourself three little RF voltmeter circuits that measure the voltage across the input, the known series R, and the 'unknown' (lets call them V1, V2 and V3).

    Initially measure some known 'unknowns' so that you can verify circuit operation.

    In the general case, the unknown can be a complex impedance, ie resistance and reactance.

    There is a relationship between the three voltages are phasors which by Kirchoff's Voltage Law form a closed triangle. By trig, you can find the magnitude and angle of the unknown Z (cos rule, sin rule), but there is ambiguity, you don't know whether it is a positive or negative angle. That ambiguity can often be resolved with knowledge of the load by measuring again at a slightly different frequency.

    There is also a graphical solution that is easy if you don't have trig.

    This technique is not practical for extremely low impedances or extremely high impedances, but you should be able to realise 10Ω to 1000Ω.

    This is how I measured antenna impedance when I was a kid at school and VNAs were out of the question.

    It is a good excercise in learning about complex impedances and RF measurements, something that will stand you in good stead if you do eventually acquire an Antenna Analyser or VNA or the like.

    The challenge of measurement is to use what you have and what you know to make a valid measurement and to understand the uncertainty.

  7. W5DXP

    W5DXP Ham Member QRZ Page

    Yes, all that you describe is "possible" but in a real-world system, of all possible conditions - the percentage of special conditions that can cause a false positive are relatively low.

    I agree with that statement, Owen. But the MFJ-835 does seem to provide a reliable indication of imbalance, i.e. if the MFJ-835 indicates imbalance, the currents are most likely unbalanced. It is the balanced indication that is questionable.
  8. VK1OD

    VK1OD Ham Member QRZ Page

    If you route BOTH wires of a two wire line through the core, the flux in the core is due to the algebraic sum of the magnetising force (Ampere turns per metre) of both conductors. The meter responds to the imbalance or common mode current.

    There are other designs around which use a split core (eg GM3SEK) which can be more convenient, but the solid core that Tom has used eliminates calibration uncertainty if the core is properly closed with each measurement.


    Just keep basic magnetic principles in mind when you wind the current transformer. The designer of the above described it as a low capacitance three turn balun, but it is a high capacitance one turn balun.

    Each pass through the core in the same direction is one turn!

    Yep, go for it, build it.

  9. KM1H

    KM1H Ham Member QRZ Page

    That reverse direction method was described by a W1 in the early 80's and debunked shortly after yet it still shows up in some antenna manuals and on-line and actually used in some commercial antennas. Poor engineering seems to survive no matter what.

  10. VK1OD

    VK1OD Ham Member QRZ Page

    W1** is not very specific.

    Joe Reisert (W1JR) did describe a balun with a crossed over winding around 1978, but in his balun, every turn went through the core in the same direction. Reisert's scheme is represented to have a higher self resonance, but in my experience, sometimes it has, sometimes it hasn't. It is certainly convenient for arranging the input and output on opposite sides of the toroid and is widely used.

    The pic I offered is from an online article describing an amplifier, and I have not seen that configuration before or since. It seemed to me to be a mistake, but I defer to you if you can give a specific reference that proposed that technique as valid.

Thread Status:
Not open for further replies.

Share This Page