From: Paul
Date: Mon, 28 Aug 2000 01:50:33 +0100
Subject: [TSSP] New results shed light on input impedance discrepancies
Results on: BIG LTR Coil 45 Inch High Toroid
http://63.225.104.218/test/TeslaCoils/MyCoils/BigCoil/BigCoil.htm
Many thanks to Terry Fritz for providing a set of measurements
on a coil set up in a well defined environment.
Using the Nagaoka inductance 76.45mH to normalise Mpt:
Measured Model Error
97.87 95.9 -2.0%
321.40 314.7 -1.9%
490.16 479.2 -2.2%
which, as expected, are a little low. Terry thoughtfully
mentioned an inductance shift due to ground plane proximity,
and using measured in-situ 74.1mH instead of Nagaoka, the model
predicts:
Measured Model Error
97.87 97.36 -0.5%
321.40 319.7 -0.5%
490.16 486.7 -0.7%
Input impedance (fiddle factor W set at 0.5):
Measured Model Error
f1 408 ohms 180 ohms -56%
interestingly, compare with other coils:
my large coil:
Measured Model Error
f1 120 ohms 68 ohms -43% my 30kW CW coil
f1 130 ohms 52 ohms -60% my small coil
suggests that W should be doubled, ie set to one,
ie removed, ie one less fiddle factor.
(W was set to 0.5 to calibrate the model to Thor's Q factor.
Maybe there is something wrong with Marco's Q measurement?
I hope there's a mistake in the stored energy calculation,
if its under-calculated by a factor of 2, then everybodys
measurements are compatible.)
Definately need more input impedance and Q factor
measurements.
With a score of 3:1 in favour of unity W, I'll prepare
v0.2d software with W=1 and recompute all the coils.
A final recompute, with W=1 and using Terry's measured 74.1mH,
Frequencies remain as above, now input impedance is
Measured Model Error
f1 408 ohms 368 ohms -10%
but the Q estimated is a miserable 57. I suspect it should be
about double this.
Terry, can you measure your Q factor?
Discrepancies:
A couple of glitches in the external cap matrix, laplace
solver aliasing problem.
Medhurst tables give 15.6pF, cap matrix says it should be
11.7pF, 25% out - its usually nearer than that. However,
our figure 11.7pF + top load 25.4pF gives correct f1
in the lumped LC model.
Data (based on final run with W=1, Lcoil=74.1mH):
See http://www.abelian.demon.co.uk/tssp-data/tfltr45/
Files are (in CSV format for Terry running windows):
tfltr45.internal-caps.csv: internal caps matrix, pF, 32 segments
tfltr45.external-caps.csv: external caps matrix, pF, 32 segments
tfltr45.vi97360.csv: VI records at f1.
tfltr45.vi97360.gif: VI plot at f1.
General info:
simulator version: v0.2d, 28th August 2000
system: tfltr45
LCR Sections 1000, Capacitance segments 32
Coil turns: 1000, resistance: 69.87
Coil radius: 0.130, length: 0.762
Conductor radius: 0.255 mm
Spacing ratio: 0.670
Top loading capacitance: 25.42pF
Weighted internal toroid cap to coil: 5.74pF
Weighted external coil cap to gnd: 8.49pF
Weighted internal coil cap: 3.16pF
Medhurst C should be: 11.65pF
Standard Ln: 1.075uH
Quoted inductance: 74.10mH
Lundin inductance: 76.45mH
using quoted inductance
chose sigma for Mpt: 1345.91
mid coil turn: 79.86uH
end coil turn: 41.10uH
Regards All.
--
Paul Nicholson,
Manchester, UK.
--
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.