How to use MFJ Antenna analyzer

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by WA2SQQ, Feb 7, 2011.

Thread Status:
Not open for further replies.
  1. WA2SQQ

    WA2SQQ XML Subscriber

    I'm wondering if:
    1. Anyone know how to use the antenna analyzer to measure the resistance (at 1.8 mhz) of a choke balun I constructed?

    1. Anyone know of any sites that discuss how to use the analyzer? I've already looked at MFJ's posts on YouTube - these are great, but I'm sure hams have posted their own tips.
     
  2. KB4QAA

    KB4QAA XML Subscriber

    Which analyzer do you have? MFJ has made at least six different models.

    Resistance?? Do you mean impedance?
     
  3. AC0FP

    AC0FP Ham Member

    I used a network analyzer to measure a ferrite bead choke on a 1' piece of coax that I made.

    Resistance measured zero.

    Impedance measured ~ 1000 ohms from 3 to 28 MHz peaking around 10k ohms at 14 MHz.

    fp
     
  4. NA7U

    NA7U Ham Member

  5. VK1OD

    VK1OD Ham Member

    As noted by someone else, your objective is to measure the impedance, it is a complex quantity with resistance and reactance.

    Join both wires at the input and at the output of the balun, and connect them between the analyser's ground connection and centre pin. There should be nothing else connected to the balun.

    Now put in in impedance measuring mode, and write down / plot the impedance components over a range of frequencies. It is most likely that the reactance component at the low frequency is positive, but at much higher frequencies, it will probably become negative.

    The impedance of a good common mode choke may be higher than you can measure with the MFJ259B for example.

    Owen
     
  6. KM1H

    KM1H Ham Member

    If those are Fair-Rite ferrites then go to their web site and look up the impedance curves over the frequencies you want. Im not aware of any MFJ instrument that will be satisfactory.

    Carl
     
  7. W5DXP

    W5DXP Ham Member

    The MFJ-259B is limited to measuring impedances less than ~650 ohms. If your choke balun is a 1:1, just short the input wires together and the output wires together, and measure the impedance. If it is less than 650 ohms, the MFJ-259B should yield a reasonable result. Hopefully, the impedance will be too high for the MFJ-259B to measure.:)
     
  8. G3TXQ

    G3TXQ Ham Member

    Do you mean that the DC resistance measured zero? It would be very unusual if a choke had zero resistance from 3 to 28MHz.

    73,
    Steve G3TXQ
     
  9. G3TXQ

    G3TXQ Ham Member

    Unless I'm mistaken, all you'll find of relevance on the Fair-Rite web site is curves of μ' and μ'' vs frequency. You need more information than that to predict the choking impedance.

    Steve G3TXQ
     
  10. WB2UAQ

    WB2UAQ Ham Member

    If the Z is above the MFJ's range, you can extend the measurement range by using other known passive devices such as a resistor across the terminals in parallel with the unkown part under test. With the impedance now within range of the MFJ unit, you can mathematically remove the known resistance revealing the higher unknown impedance of the part in question. My homebrewed impedance measuring instruments and some of my ancient commercial instruments are limited in the same way so if I can do it, anyone can:) Plus it is a good exercise in learning AC circuit theory:) For example, my old noise bridge (applicable for a Palomar as well) can only reach an Rp (parallel equivalent resistance) of 250 Ohms so I would place a 330 Ohm low reactance resistance in parallel to the unknown and null the bridge under this condition and then remove the 330 ohms mathematically. If any one would like to work thru an example, let me know and I'll go thru it step by step. The useful formulas here are the series to parallel and parallel to series conversions. I programmed these into my HP calc and have also loaded everything into a spreadsheet. 73, Pete
     
  11. G3TXQ

    G3TXQ Ham Member

    Pete,

    That method's useful, but we need to recognise that the accuracy of the measurement deteriorates.

    For example, say we have a choke whose impedance is 2000+j2000 and we shunt it with a 330 Ohm resistor; the parallel combination is 303+j23. Now let's suppose the analyser reads 10% low on Resistance and Reactance, showing us 273+j21. We will deduce that the choke impedance is 1350+j620 instead of 2000+j2000.

    But if it's the only tool we have, then "needs must"!

    73,
    Steve G3TXQ
     
  12. VK1OD

    VK1OD Ham Member

    There is a whole lot of mixed advice here.

    I have explained how you would use an analyser to make such a measurement.

    All instruments are limited in their 'accurate' range, and many inexpensive equipments will be challenged by a good common mode choke.

    Most general purpose HF common mode chokes are shy on choking impedance at 160m, and the real choke might be within range of even low end instruments at that frequency.

    Anyone with practical experience of broadband ferrite baluns knows that many if not most designs will be self resonant within their broadband range, and the effects of stray capacitance need to be added in to any theoretical prediction. Like Steve, my experience with the Fair-rite site is that curves of complex permeability vs frequency are publised for many cores, that is not an "impedance curve", and there are "impedance" figures published for single turn windings on a small number for cores, primarily those for suppression applications. Single turn figures cannot simply by multiplied by turns square for higher turns.

    The accuracy of low end instruments is not good, and the thought that an instrument and fixture with a limit of perhaps 600 ohms could be extended to several k with any accuracy is naive.

    As always, it is useful to measure some knowns with an instrument to become familiar with its behaviour. The following experiment will not only improve familiairity with the instrument, but demonstrate some of the effect expected of the ferrite, stray capacitance and range issues in measurement.


    1. Take a medium µ ferrite toroid, say #43, of medium size.
    2. Place one turn on the core (that means the wire passed once through the centre, there are no half turns on a toroid.
    3. Connect it between ground and centre pin of the instrument.
    4. Measure, tabulate and plot R and X from 2 to 50MHz.
    5. Now, repeat this for two, four and eight turns.
    Note that in doubling the turns from one to two, impedance (R, X) increases by approximately 4 times. You might observe a similar effect from two to four turns.

    At some point, you should observe the effect of stray capacitance and the increased inductance give rise to a resonance within the measurement range. At resonance, X will pass through zero and change sign though some instruments do no directly indicate the sign of X.

    So, you have found resonance, bunch the turns and compare it with the turns spread over 270° of the core... what does it do for resonance and impedance above resonance? What if the winding is distributed to use 360° of the core, what about the W1JR cross over configuration. What is the impedance at or above resonance, is it still a useful choke?

    Remember that an instrument limited to say 600 ohms or so is not going to reveal some of these effects, but worth the exercise for what you will learn.

    Now you are learning, learning more than someone telling you that it is all in graphs on Fair-rite's site.

    [​IMG]

    Above is a graph of what you might expect with sufficient turns. In this case it is 12t on an FT114-43. The graphs is from a model that was calibrated to measured impedance (using a VNA) from the 12t inductor.

    Readers with a VNA can perform the same measurements, it is just quicker, gives graphical output, and most VNA's will have sufficient measurement range to capture impedance of a practical common mode choke.

    Owen

    PS: This might wet your appetite to do the same test for a powdered iron core. Do they make good general purpose common mode chokes? Why?
     
  13. WB2UAQ

    WB2UAQ Ham Member

    Owen,

    Steve pointed out something quite some time ago in regard to hi Q material such as iron powder. When in a multiband common mode choke application they can create a sharp series resonance condition making the system worse off with the balun at some frequency. Maybe this is where you are going?

    I can't argue with the accuracy issues and the lack of resolution, etc.. I have to deal component measurements at the salt mine and often corellation from one measurement technique to another is put to the test. For instance when the impedance of a device gets to far from 50 ohms a VNA might not be the best instrument either. I have an old copy of HP's Impedance Measurement Handbook (likely the newer version can be optained from Agilent's website). There is an enlightening chart comparing various measurement techniques comparing bridge type, 4 wire type and reflection type measurements and maybe more.

    73, Pete
     
  14. VK1OD

    VK1OD Ham Member

    Hi Pete,

    I cannot recall Steve commenting on series resonance of the inductors. Of the many that I have constructed for HF and measured, they usually exhibit a parallel resonance, but have not shown clear series resonance in that range.

    Steve has commented on so called Ugly Baluns which usually show a high Q parallel resonance, and as you would expect with a high Q inductor, only modest impedance well away from resonance, and quite often ineffective in common mode suppression. It is clear that Ugly Baluns do not provide broadband high choking impedance, and are not well suited to that application. I have used them recently, but always in a single band application and they were tuned to peak in the band of interest.

    It is my experience that iron powder cored inductors are not quite as sharp as the Ugly Balun, but the same issue exists, mediocre choking impedance. If more turns are piled onto a core to increase inductance, the increased self capacitance reduces the frequency of resonance, and impedance above resonance quickly falls to low magnitude. I know these cores were once very popular for antenna common mode chokes, but more informed thinking relegates them to second rate.

    Yes, RF impedance measurement is challenging, and hence my reservations that an instrument with limited accuracy cannot be extended to say 5 times range with much accuracy at all. We learned those lessons with the extension technique use on inexpensive noise bridges if you recall.

    Owen
     
  15. G3TXQ

    G3TXQ Ham Member

    Owen,

    It may be that Pete is remembering comments I have made about Hi-Q chokes being predominantly Reactive away from their resonant frequency, and how that Reactance can actually increase CM current if it is of opposite sign and similar magnitude to the CM path impedance.

    73,
    Steve G3TXQ
     
  16. WB2UAQ

    WB2UAQ Ham Member

    Steve and Owen
    That is exactly it. If the common mode impedance is capacitive, it is possible to resonate with the high Q inductance of the iron powder-cored choke creating a low Z path. At one time I stated that it is only important for the choking impedance to be high regardless of its nature overlooking the case of the "opposite sign and similar magnitude" as Steve states above.
    Pete
     
  17. AC0FP

    AC0FP Ham Member

    What I said, if the resistance is less than 1 ohm its zero in my book!
     
  18. VK1OD

    VK1OD Ham Member

    So, if I put those together, R=0, and Z varies with freq to a max of 10,000. If Z is 10000, and R is approx zero, lets say it was actually 1, then X is 10000 and Q is X/R=10000... very unlikely.

    If this was an air cored inductor, a Q at HF above 500 is quite high, for a powdered iron cored inductor Q above 200 is high, and for ferrite, Q where Z peaks is quite likely to be single digit.

    Something just doesn't gel.

    Was this a single core, like a suppression sleeve, or many of them, was it ferrite?

    Owen
     
    Last edited: Mar 14, 2011
  19. AC0FP

    AC0FP Ham Member

    This measurement was taken sometime ago, so memory is a little hazy. The DC resistance was <1 ohm. I really didn't note the real part measured by the network analyzer as I was just interested in the impedance (Z).

    The choke was made up of 15 ferrite beads slipped over some RG58C/U coax.

    fp
     
  20. VK1OD

    VK1OD Ham Member

    DC resistance would usually be very small and irrelevant.

    Yes, some people are interested only in the magnitude of impedance, others understand the value of both magnitude and phase, or R and X components... and the R component for ferrite cored chokes is usually not insignificant.

    Owen
     
Thread Status:
Not open for further replies.

Share This Page