Ferrite for baluns

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by KC0LWN, Feb 8, 2015.

Thread Status:
Not open for further replies.
ad: L-HROutlet
ad: l-rl
ad: abrind-2
ad: l-BCInc
ad: QSOToday-1
ad: Left-2
ad: l-gcopper
ad: L-MFJ
ad: Left-3
  1. E73M

    E73M Ham Member QRZ Page

    Steve, I have 2.75" spacing of two so239 connectors with only 1.5" lenght to VNA, I do thru, thru match and crosstalk calibration.

    I have higher resonant point then your measurement which point me to conclude that my jig capacitance is small.
    Now, I have added two 21" cables of RG400 with crocodile clips at the center pin and calibrated it same way. Leads are separated by 3 or more inches.
    Reading is 11MHz as resonant point and peak at 10000 Ohms just as your published curve here at QRZ.com.

    So, now I have tested another choke with #31 material which has Fres. at 2.7MHz with my jig and 2.4MHz with cables (I guess close enough setup as your jig).
    It is clear to me that I do NOT add parallel capacitance since resonance is shifting upward.

    I have tested my chokes on R&S FSH3 VNA and was getting same results and frq. shift if I cut leads shorter.

    Anyway, thank you for your time and input.

    73 Danny E73M
  2. G3TXQ

    G3TXQ Ham Member QRZ Page


    If you make an S21 measurement at, say, 15MHz with no choke connected to the jig, what do you get for |S21| and S21 Theta?

    Steve G3TXQ
  3. E73M

    E73M Ham Member QRZ Page


    My jig works perfect.
    I am still trying to figure out why your style of jig with cables gives me a huge error in measurement.
    Can you measure one turn with your jig and get correct point of u' and u" for few materials ie #31,43,52 and 61? I tried with cables but results are way of and are shifted downward.

    73 Danny E73M
  4. G3TXQ

    G3TXQ Ham Member QRZ Page


    If you mean the u'=u'' "cross-over" frequencies - yes they come out about right given the sort of tolerances quoted for ferrite materials. For example I just measured Rs=Xs at 48MHz for one of my type 61 cores, compared to around 42MHz read from the Fair-Rite complex permeability chart.

    I used cables because I wanted to keep the DUT and the jig well clear of the metalwork of the analyser. With that method you do of course have to cal out the cables and, if you use a 50 Ohm termination at the jig as I do, you need to take account of it when deriving Rs and Xs from S21 amplitude and phase.

    One extreme test for the measurement technique is to check the resonant frequency of a high-Q trap and compare it with what you get using a "no-connection" method. When I do that I typically see a downward shift with the VNA equivalent to the addition of 0.2pF to 0.3pF of extra parallel capacitance.

    Steve G3TXQ
  5. E73M

    E73M Ham Member QRZ Page


    My jig gives me a almost perfect u'=u" "cross-over" freq. for each Fair-Rite material. For example I get 38.6MHz for #61 with my jig but with cables it drops down to 15.6MHz, will look into this problem later.

    See attached my #52 and #K u'=u" measurements, you and Cecil did not show the "cross" of Rs/Xs in your measurements, also Owen (ex VK1OD) commented on his blog that your diagrams show low impedance but I think that he missed the fact that Fair-Rite posted those measurements done on small 19mm cores which will produce more CMI then 61mm (2.4inch) cores.

    Anyway here is my data for all materials I had on hand to measure witgh ONE turn on my jig, they are very accurate compared to mfg. data
    1Turn-FT240-52-web.png 1Turn-FT240-K.png 1Turn-FT240-43-web.png 1Turn-FT240-31-web.png 1Turn-FT240-43-web.png 1Turn-FT240-61-web.png

    73 Danny E73M

    NOTE: diagram without Fair-Rite data is #K material
    Last edited: Mar 20, 2015
  6. G3TXQ

    G3TXQ Ham Member QRZ Page


    Just to be clear - none of my earlier plots were derived from S21 measurements on the VNA2180! I used S11 measurements on an AIM4170 for consistency with what Cecil had done. And in fact, once he and I had unravelled some of the issues, I discovered that a trace on an early chart I posted was of his data not mine :eek:

    There has to be some major problem if you observe a drop from 38.6Mhz to 15.6MHz by adding cables between the VNA and the jig. The cables simply extend the 50 Ohm A and B ports to a different measurement plane; so provided you recalibrate to take account of the extra phase delay and the small extra attenuation which is introduced, you should see exactly the same results.

    I'm quite surprised you see such close agreement with the Fair-Rite charts, given their quoted tolerance figures and the variations that I measure between different cores of the same material; maybe you just got lucky!

    I don't think you and I differ much on this topic; but I do still maintain that if you see SRF shifts when extending the length of the choke "pigtails", it can't be due to the added series inductance :)

    Steve G3TXQ
  7. AI6KX

    AI6KX Ham Member QRZ Page

    I don't have a source of toroids here in Japan and want to order some from the US. In order to minimize shipping costs and mistakes, could someone please cut to the chase for a theory-deprived guy, and suggest a good source and appropriate types/sizes for use in baluns, potentially from 7 to 30 MHz? Power will be less than 200 watts. Many thanks!

    Steve in Okinawa ex-K2RDP and W6BMZ
  8. W5DXP

    W5DXP Ham Member QRZ Page

  9. KM1H

    KM1H Ham Member QRZ Page

Thread Status:
Not open for further replies.

Share This Page