Jim Brown wrote: >On Thu, 26 Aug 2010 08:08:16 +0100, Ian White GM3SEK wrote: > >>However, the AIM4170 is a higher-level instrument that is capable of >>being re-calibrated to include the effects of an attached test fixture >>(which then becomes "part of the instrument"). With care, it seems that >>the 4170 can be used to make quite accurate vector impedance >>measurements on chokes. The next step, having calibrated the instrument, > >Yes, the AIM is a high quality instrument, but one tool does not fit all >needs. It is, in essence, measuring S11, which means that very small >errors in the measurement can cause very large errors in the result. The >only GOOD way to measure a choke is by measuring S21 -- that is, as the >series element of a voltage divider. >
I measured the same chokes in both types of test jig, reflection and transmission, and neither method has any clear advantage over the other. Both methods have potential problems with variations in series inductance and shunt capacitance (the latter in parallel with the choke). In both cases, everything depends on the care taken to maintain the test jig in exactly the same geometry, first for calibration and then for all subsequent measurements. Among other things, this requires the construction of special reference standards which allow the instrument to be calibrated without changing the geometry of the test jig. >The errors I have seen in choke measurements show up as lead capacitance >that is not precisely cancelled, which results in an incorrect >determination of the resonant frequency. The error is small for low >frequency, low-Q chokes, but can get very large at increasing frequency, >and for higher Q chokes. > As I said previously, any choke that is liable to this kind of measurement error will have exactly the same problems in a practical installation - only far, far worse - resulting in unreliable performance. That is a very good reason to avoid high-Q (frequency selective) chokes and choose a more broadband design. >In the materials on my website, there are graphs showing the effect of >the 0.4 pF stray capacitance of my own test fixture on a very low Q choke >(Q=0.4) wound to cover a tri-band beam (20-10M). It shifts the resonant >frequency about 25%. >The error is insignificant for #31 chokes wound for >40M and below. > >The error becomes very bad with a higher Q choke, like one wound with >Fair-Rite #61 or #67 (Q around 10). That is correct: on 14MHz and above (and especially at VHF), even a tiny change in shunt capacitance can move the resonant frequency by a large amount. That creates an "interesting" measurement problem - but much more important is the effect of uncontrolled stray capacitance when that same choke is installed on the antenna. In practical installations, it means that we cannot rely on resonance alone to provide a large choking impedance, because the resonant frequency of the choke will move when the choke is installed. >The AIM data I've seen shifted the >resonant frequency of a choke like that by a factor of nearly 2:1. The >AIM measured the resonance of choke as being around 16 MHz that was >actually resonant above 30 MHz. And, because the resonance moved and the >loss of the material varies with frequency, the peak in the choking >impedance also was in error, but by a smaller amount. > There is no reason for the AIM to have that problem. Either the user was allowing the test setup to vary, or the AIM had never been correctly calibrated with the jig in place. -- 73 from Ian GM3SEK http://www.ifwtech.co.uk/g3sek ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[email protected] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html
