Several issues here. See comments inline.
On 12/31/2017 3:12 PM, Avery Davis wrote:
Maximo,
Just measure the common mode choke just like a simple inductor, i.e.,
ignore the center conductor of the coax with which it is wound.
OK. BUT -- it's NOT an inductor. See below. Also study
k9yc.com/RFI-Ham.pdf Or study it in the ARRL Handbook. It was added in
2011.
Set the measurement to S11,
Bad idea. Z measurements using S11 become increasing inaccurate for Z
that differs by more than a factor of 5 from the system impedance (that
is, less than 10 ohms and more than 250 ohms). The design goal for
common mode chokes is 5K ohms, and the equivalent circuit is parallel R,
L, and C. C is interwinding capacitance, and is in the range of 1-8 pF
for practical chokes. Because of how Z is computed from S11, very small
errors in S11 result in very large errors in Z, where Z is complex.
This, combined with stray C of a measurement setup, can result in very
large measurement errors in both the magnitude of the resonance and the
frequency of the resonance.
This limitation of S11 measurement of Z is well known. It's been in
Hewlett Packard app notes for decades, and it's why they made a
dedicated analyzer to measure Z. The method using S21 is good for high
values of Z (that is, >> 50 ohms).
A different method must be used for low Z values. In the RF world it's
called RFIV. In pro audio, a method was developed whereby a large series
resistance was added to the output of the generator to make it a
constant current source, and the voltage across the load is measured by
high impedance voltmeter. This method is widely used to measure Z for
loudspeakers, using a simple swept oscillator and audio voltmeter, or
Time Delay Spectrometry, or dual channel FFT for instrumentation. A
current limiting resistor of 5K ohms yields 10% accuracy for Z mag at
500 ohms and 1% at 5 ohms.
73, Jim K9YC
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|