From: Paul
Date: Thu, 07 Jun 2001 23:22:34 +0100
Subject: Re: [TSSP] Interesting article on Medhurst, Wheeler, modeling,etc.
Finn wrote:
> A close count revealed that I had 505 turns. Stripped off those 5
> turns, the coil now measures 144.5 mm long, and 28.8 ~ 28.9 mH.
> (1kHz out of the Isotech LCR.)
> DCR is measured to 87 ohms.
That adds up now. Amazing how you pack the wire so well on that coil, it
must be very nicely built. With your revised dimensions, the figures
are now:
fh1: bare d=0.160m h/d=0.90 sr=0.86 b/h=6.92 turns=500
measured modeled
f1 359.5kHz 360.4kHz +0.3%
f3 839.2kHz 883.7kHz +5.3%
> Material: 5.5 mm thick PVC
That's pretty thick. Consider: at f9 the significant length over which
Cint applies its (purported) effect most strongly is around 32 mm, which
is the length spanned by a half wave (across which twice the terminal
voltage is induced). Half the Cint flux (that which goes inside the
coil) must pass through at least 2*5.5mm material, which amounts to
25% of the minimum path length, and most of the flux joining two points
a half wavelength apart would be expected to take rather more length
inside the material. If the relative permittivity of the material was
2 or so, then it's easy to see how the Cint over a half-wave could be
increased by 50% or more, resulting in a 20% or more reduction in
frequency of f9.
For this reason, I'm encouraged to consider modeling the effect of the
former material. Perhaps this may bring the small radius results into
line with predictions. We have quite a few of these!
> hard to adjust precisely enough to get a stable lissajous pattern,
> so i skipped the higher modes for now)
Pity, they are the most informative. The pattern is too accurate for
this then. We only need 0.1% at best, that's a 1kHz rotation of the
figure. Your eye is forcing you to tune to 0.001% or better! Couple
both oscillators through a diode into an audio amplifier, tune till you
hear the beat go below 1kHz.
> Are data from these small coils valuable? I would rather move up to
> larger coils with frequencies that I can count.
Larger coils are better - less sensitive to material thickness, less
sensitive to probes, lower frequency so the overtones are easier to
reach, yep, larger coils. But I think we can do business with the one
you have for the moment - it's just about big enough.
> You mentioned that you need time to code,
Give me 36 hrs/day and I'll still want more,
> so perhaps it would be a good idea that you make a wish list of coils
> that you want data from, then I can make them and we can pick up the
> thread when I have gotten the generator and counter stuff sorted out,
> and you have got the code sorted out?
Yes, what I'll do is bring the things-to-do list up to date. I've a mind
to take a couple of days off work to try and crack this Cint thing. It's
such a fundamental point and so important to the understanding of the
coil above f1, we really have to be absolutely sure about it.
Cheers,
--
Paul Nicholson,
Manchester, UK.
--
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.