I tried a few more measurements this morning. I wound a choke on a
FT240-61 core and measured its SRF using various methods.
* I used the test set-up shown in Fig 40 of Jim's paper, but included a
10k resistor in series with the choke to try to reduce the effect of any
shunt C from the source. Resonant frequency was 23MHz.
* Next I tacked a 3.3pF capacitor across the choke. The SRF dropped to
11 MHz, from which I estimated that the choke equivalent parallel
capacitance was about 1pF
* Then I removed the capacitor and the 10k resistor so that I had the
exact same set-up as Jim's Fig 40. The SRF was 18.8 MHz, suggesting that
the stray capacitance of the test fixture was 0.5pF - similar to Jim's
quoted 0.4pF
* Finally I measured the SRF using the AIM; the result was 17.7 MHz.
That suggests the AIM error was the equivalent of 0.7pF of stray
capacitance.
So, if we take the "true" SRF as 23 MHz, Jim's measurement method
introduces an error in SRF of -18% and the AIM an error of -23%
It all seems pretty academic when I could shift the SRF by similar
amounts simply by changing the way the leads were dressed. As Ian said
earlier, a choke which achieves high resistive impedances is the only
way of ensuring a robust and repeatable design.
73,
Steve G3TXQ
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|