From: Paul
Date: Wed, 21 Feb 2001 12:29:36 +0000
Subject: [TSSP] Problems with high elevation/small radius
Hi All As you may be aware, simulations have been singularly unable to repro- duce f1 measurements on coils at high elevation, hence the restriction of the model to b < h. For example, we have mz1: bare d=0.051m h/d=6.00 sr=0.92 b/h=2.33 turns=875 f1 885.0kHz 1048.9kHz +18.5% f3 2338.0kHz 2525.1kHz +8.0% f5 3436.0kHz 3547.5kHz +3.2% mz2: bare d=0.051m h/d=6.00 sr=0.87 b/h=2.33 turns=1310 f1 645.0kHz 705.0kHz +9.3% f3 1627.0kHz 1697.1kHz +4.3% mz3012-5: bare d=0.089m h/d=3.18 sr=0.88 b/h=1.97 turns=622 f1 665.9kHz 721.4kHz +8.3% f3 1591.1kHz 1694.4kHz +6.5% f5 2277.9kHz 2399.1kHz +5.3% from Mark Rzeszotarski's measurements. Up to now I've been assuming that the problem with these is poor determination of the external field - the pattern of reduced errors at the higher overtones supports that. However, I've been unable to reproduce the measured f1s using reasonable guesses at the likely extra unaccounted contributions to Cext. We also have some low elevation results which exhibit the same kind of error, eg from Mark, mz3012-1: bare d=0.089m h/d=3.18 sr=0.88 b/h=0.07 turns=622 f1 647.8kHz 696.2kHz +7.5% f3 1575.4kHz 1656.0kHz +5.1% f5 2264.1kHz 2369.4kHz +4.7% and more recently from Kurt Schraner (welcome to the tssp list Kurt!), sk5b185: bare d=0.051m h/d=8.03 sr=0.91 b/h=0.45 turns=934 f1 919.5kHz 1027.0kHz +11.7% Both mz3012-1 and sk5b185 are at low elevation, albeit over poorly defined grounds, so I want to explore the possibility that it is the small radii that is to blame, rather than high elevation, in the hope that we can clear up this long standing problem. The only other narrow coil that we have a result for is Marc Metlicka's coil #1, ie mm1: bare d=0.091m h/d=8.92 sr=0.76 b/h=0.41 turns=1221 f1 455.5kHz 466.3kHz +2.4% which also begins to show signs of a positive prediction error. The 'small radii' angle opens up a few more options, eg former material relative permittivity is one that comes to mind. Up to now the working assumption has been that the volume of material is small compared to the coil diameter and has negligible effect. This may not be true at small radius. Right now I've no means of calculating the effect of former dielectric on the Cint and Cext physical capacitances. Does anyone have a small coil, eg < 8cm diameter, of thin former material, into which another tube can be slid inside, to a reasonably close fit? The with/without f1 readings would give us an idea whether an f1 shift of the right order of magnitude can possibly be caused by the former material. Of course, another possibility is that small amounts of stray capacitance, unaccounted by the program, are depressing the measured frequencies. We might expect this to affect both small radii and high elevation coils. An additional 1pF near the top of mz3012-1 and sk5b185 is enough to cure the problem, but I can't see where this might be coming from. Suggestions, and further measurements on small radius coils, are most welcome. Cheers, -- Paul Nicholson, Manchester, UK. --
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.