Steve Hunt wrote:
> Jim,
>
> Instruments like the AIM attempt to "calibrate out" the fixture. You
> show the instrument a short-circuit, an open-circuit, and a known load,
> right at the measurement plane. It can then remove the effects of the
> fixture for unknown impedances connected at the measurement plane. I
> maintain a set of calibration files for my various test fixtures.
OK. just like any modern VNA.. and that's fairly straightforward when
you're talking about a coax connector. But when you're measuring
something like a RF common mode choke or a RF transformer/Balun, which
essentially is a leaded component, what do you use as your cal standard,
particularly for the "open"?
I was trying to measure the properties of a trap that is essentially in
series with a 2 foot length of aluminum tube. Replacing it with a
untrapped piece of tube isn't really a short. Nor is hooking a 50 ohm
resistor between the plates of the fixture really a load. Not when
you're looking at 10 meter wavelengths, so a test fixture that's almost
a tenth of a wavelength long is significant.
What people who do it for a living do is use cal standards that have
properties that are known by calculation from mechanical properties
(e.g. a coax airline) that can fit in the fixture. Then, they use a
customized calibration algorithm that can use that value. I suspect
nobody is doing this with the AIM, nor does the calibration algorithm
allow cal standards other than the usual S,O,L,T.
The other issue is the basic way many of these things actually do the
measurement. For instance, they have a source and load impedance of 50
ohms, and all the levels and sensitivities of the detector are set up so
that 50 ohm-ish Zs put the voltages at "good" places in the operating
curve of the detector.
Hooking a 5000 ohm resistor in a circuit intended to measure 50 ohms
means the voltages are going to be a factor of 100 from what the circuit
was designed for. That might not be a problem, but it also might.
>
> As I said earlier I tried measuring a 10k resistor ; I also just tried
> measuring a 10k resistor in parallel with a 3.3pF capacitor. Over the
> range 1-50MHz the results seemed plenty accurate enough for making CM
> impedance measurements on chokes.
>
> I'm still puzzled why Jim (K9YC) was questioning the accuracy of the
> Balun Designs impedance data. The chart shown here:
> http://www.balundesigns.com/catalog/1-1%205kw%20Choking%20Impedance.jpg
> doesn't seem to contain impedances that would cause significant errors
> on the AIM.
>
> 73,
> Steve G3TXQ
>
>
>
> jimlux wrote:
>> Steve,
>> what about a high reactive impedance.
>>
>> Mostly, the problems I've had with this kind of thing have to do with
>> fixturing and/or calibrating the fixture out. I was trying to measure
>> the Z of traps, and never did get a good fixture scheme worked out.
>> It's not like you can just take a couple banana plug clipleads.. I
>> finally decided to remain ignorant of the electrical properties of the
>> traps in a 6-BTV.
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|