From: Paul
Date: Tue, 08 Oct 2002 08:29:48 +0100
Subject: Re: [TSSP] Curiouser and curiouser!
Terry,
Your Q factors have just about doubled, comparing now with
Sept 14th,
Sep 14 Oct 7
f1 37.801 kHz 37.229 kHz
f3 113.550 kHz 113.864 kHz
f5 171.773 kHz 172.448 kHz
q1 42.94 84.19
q3 19.45 46.26
q5 14.32 34.37
Not much change to the frequencies, f1 down a percent an the
other two up 0.3% and 0.4%. Greater Q improvement for the
higher modes, so whatever the origin of the loss that you've
lost, it was either coupled reactively to the coil, or
concentrated in one region. What a difference being indoors
for three weeks makes. Have you altered ground arrangements?
I guess not. I recall similar variations with your big LTR
coil.
When you've built your plastic cored high Q replacement,
see if you can ping it with and without a length of sonotube
inserted down the middle. I guess you'd need two lengths
of sonotube, one bone dry, the other allow to become damp.
That would confirm once and for all the suspicion about
sonotube.
I don't know what's happening to your 3/4 wave mode on
the running coil. It seems to be breaking up and tcma
is making a hash of it. Why isn't the 1/4 wave mode split?
Did you capture the ping starting after the quench? Or is
your primary desperately out of tune? Maybe the f3 of
the sec is being modulated by alternative f1 half cycles,
with the IGBT's switching on and off. 114kHz when the IGBT
is off, and 120 kHz when the IGBT is on (or maybe the other
way?), the modulation occuring due to the pri-sec cap and
the effective Lsec being altered.
You could test for that by pinging with the pinger and
looking for the change in f3 as you manually make/break
the primary loop. If such is the case, why aren't f5
and f7 upset? Most curious.
Seeing that the 3/4 wave overtone of your loaded secondary
is only a little above the 3rd harmonic of the 1/4 wave,
perhaps the rectification in the primary circuit is
generating enough 3rd harmonic of f1 to upset tcma. But
I'd have expected something at 111.4 kHz. Still, it would
explain why the other modes aren't affected.
(from pn2511, eq 9.8)
R = 2 * pi * F * Lee/Q = 2 * pi * 37229 * 0.416/84.2
= 1200 ohms.
That's the coil resistance referred to the base terminal,
ie the input R.
tssp predicts 78 for your q1, based on the estimated AC
resistance of the winding plus an estimate of proximity loss
from Medhurst tables. Thus I wouldn't expect your Q to rise
much more, ie your sonotube must be pretty well dried out by
now.
--
Paul Nicholson
--
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.