[TowerTalk] Test Fixture for Common Mode Chokes

[Brian Beezley]

AB7E wrote:

"Any thoughts on this text fixture?"

Dave, I completely forgot that I had contacted Halibut Electronics, the 
fixture manufacturer, a couple years ago. Mark Smith sent me two .s2p 
files and photos of chokes he had measured through the fixture.

One of the chokes had two Fair-Rite mix 31 ferrite tubes (not cores) in 
series. Each was separately wound with three loose turns of "ultra flex" 
RG-8X. The S21 and Y21 |Z| plots were not that close and both had funny 
wiggles (the same was true of the R plots). Shunt capacitance was 6-7 pF 
(but much larger below 20 MHz). Dissipation factor (DF) was very high at 
30-50%.

I do not think these results are characteristic of the fixture. They are 
inconsistent with those of the second choke, which was 13 turns of RG316 
on an FT140-43 core:

https://iili.io/fQHD64j.png

|Z| was essentially the same for the S21 and Y21 methods, as was R (30 
MHz max). Except for noise at lower frequencies, shunt capacitance was 
very low:

https://iili.io/fQJB4Mx.png

DF was between 10% and 20% above 15 MHz, but much higher below 7 MHz. 
These are unusually high values (a few percent is typical). Where did 
the loss come from? The NanoVNA V2 Plus4 Mark used did not generate S12 
and S22. The analysis set S12 = S21 and S22 = S11. This is reasonable 
for a symmetrical device and fixture. Mark said he placed the choke in 
as much empty air as possible for the test. He used the calibration 
standards in the fixture.

Other than the odd DF values, I'd say the fixture looks fine.

Brian