From: "Malcolm Watts"
Date: Fri, 23 Feb 2001 08:39:48 +1300
Subject: Re: [TSSP] Problems with high elevation/small radius
Hi Paul, all,
Again, the fact that a piece of wire sitting above the
top of the coil adds significant capacitance is well known. There are
corrections for this cited by Medhurst and others and I have seen it
occurring in old test. It is a remarkable fact that for a particular
coil geometry, a piece of wire as long as the coil itself and sitting
bolt upright on top (and obviously connected) can double the
effective Cself of the structure.
Regards,
Malcolm
On 22 Feb 01, at 7:59, Paul wrote:
> Marc wrote:
>
> > the wire is 2.5" long from top turns and curls about 1/4
> > turn to the bolt.
>
> Right, so no significant additional inductance. I just did
> some sums on that wire length, and the calculated C for 2.5"
> of wire at that height above a groundplane is... 0.45pF. Nice
> measurement Marc. The total 'end-effect' capacitance for this
> coil is around 1.1 pF, so even though the additional wire is
> only 20% the length of the coil's circumference, it adds another
> 50% to the end capacitance, I guess because, unlike the top turn, it
> doesn't have the sheilding effect of the coil beneath it.
>
> Terry wrote:
>
> > I noticed too while working on E-Tesla that either having
> > a top center connection or just a solenoid without and end
> > wire makes a significant difference in f1. If a coil's diameter is
> > large, there is a big air space at the top. Having a wire leading
> > to that space can add a bunch of top capacitance.
>
> Yes, I've just been neglecting these little things. In fact I
> recall, during discussions on E-Tesla, saying that you could
> surely ignore those bits - well I guess I was wrong on that one!
> Marc's little observation has made that clear to me, especially on
> small coils, we'll have to take account of any additional wires
> extending from the top of the coil.
>
> Perhaps Kurt and Mark can let me know if there are any lead wires or
> terminals coming off the top turn?
>
> Marc wrote:
>
> > i think i'll do a full battery of tests on #2 and #3.
>
> No harm in repeating f1,3,5 on #2,3. There is always some
> variation each time you measure the coil. Inevitably there will
> be quite some scatter in the readings, eg, here are some f1
> comparisons from my database of measurements:
>
> measured modeled error
> f1 150.7kHz 150.2kHz -0.3%
> f1 148.4kHz 146.5kHz -1.2%
> f1 455.5kHz 466.3kHz +2.4%
> f1 276.9kHz 273.8kHz -1.1%
> f1 61.9kHz 63.7kHz +2.9%
> f1 237.0kHz 243.0kHz +2.5%
> f1 405.1kHz 408.8kHz +0.9%
> f1 217.2kHz 208.0kHz -4.2%
> f1 161.4kHz 156.4kHz -3.1%
> f1 358.8kHz 357.4kHz -0.4%
> f1 90.9kHz 91.3kHz +0.5%
> f1 224.0kHz 225.5kHz +0.7%
> f1 220.0kHz 225.4kHz +2.5%
> f1 65.5kHz 66.5kHz +1.6%
> f1 97.9kHz 95.3kHz -2.6%
> f1 66.7kHz 65.9kHz -1.1%
>
> While the average error is around zero, as you can see there is a
> fairly wide distribution of errors around the mean. All sorts of
> factors are likely to be involved, temperature, humidity maybe,
> position. This kind of scatter suggests that we can only get below 1%
> by considering an average of results across several coils, and perhaps
> by using an average of several readings taken on each coil, taken on
> different occasions.
>
> Cheers,
> --
> Paul Nicholson,
> Manchester, UK.
> --
>
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.