[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Secondary Q variation experiment is going!!



Original poster: "Paul Nicholson by way of Terry Fritz <twftesla-at-qwest-dot-net>" <paul-at-abelian.demon.co.uk>

Hi Terry, All,

Thanks for the neat summary of the experiment.  I'll just follow
it up with some observations on the data collected so far.

This first run 'expt0' is just a cheap and cheerful test run to
see that it's all working.  We're finding that the experiment is
reliable and stable, and that the coil is delivering just the kind
of Q and F variations that we are looking to investigate.

Referring to 

http://hot-streamer-dot-com/bob/qvar/results/expt0/index.html

This web page, which is automatically updated roughly every hour,
shows the progress of 'expt0'.

Consider the second to last graph - the Q-factor trend with
humidity.  The coil was showing a nice high Q factor until
04:00 hours on the 8th, when it fell by a huge amount (from 163
down to 28) over about 30 minutes.  This coincided with a rain
shower.  The coil gradually recovered over the morning and was back
almost to its previous level by midday.   A similar event occured
the next night, no rain this time, but the temperature fell to the
dew point (ie humidity reached 100%) and we see a steady ramp down
of the Q over a period of a couple of hours.   

First impressions are that the Q is more or less independent of
humidity until the point it reached where water starts to condense
out (rain or dew), and we believe the roof tiles are becoming damp
enough to contribute a significant loss.  A dry day 'hosepipe' test
is planned to confirm this.

Associated with these wild Q drops is a modest but significant
frequency drop, as seen in the 5th graph.  The frequency fell by
0.4% coincident with the rain shower on the 8th.  This provides
strong evidence that the additional loss which is decimating the
Q factor is coupling to the resonator via its E-field rather than
the B-field.  If the loss were coupling via the B-field (ie eddy
loops forming in the damp roof) we would see the frequency rise
slightly, not fall.

Jim wrote:
> The resistance of the secondary will change significantly with
> temperature..  does that correlate with observed Q at all?

Yes!  We are seeing both Q and F variations which correlate with
temperature, although with the coil in its present position it is
difficult to factor out the grosser variations due to moisture.
One thing that helps us is that variations with temperature are
expected to affect all three mode's F and Q by the same percentage,
and if we ignore the wild excursions discussed above, we do indeed
see the F and Q of all three modes gradually varying in unison.

The 4th graph - showing Q-factor as a scatter against temperature
allows us to estimate the background variation. Putting a line of
regression (aka a ruler!) through all the points gives us around
-0.8% change in Q per degree C rise in temperature, for all three
modes.  That's about twice what we would expect based on copper 
resistance alone, but we can't draw any conclusions from that
because in this particular run the ground current return path is
uncontrolled.

We also see signs of a nice change in F with temperature. This is
demonstrated very well in the 1st graph from 9th 08:00 hrs to 20:00
hrs when the temperature went through a strong daily peak. All three
mode frequencies fell by around 0.2% with the 12 deg C variation in
temperature, and the regression line has a slope of around -0.02%
per deg C.  Presumably we are seeing the frequencies drop due to a
thermal expansion of the coil.  It might be interesting to try this
with an unvarnished coil so that the expansion is unconstrained.

We will no doubt set up another run at some stage to bring out more
clearly the effects of temperature, but first we have to understand,
and then dispose of, the grosser variations due to moisture.

One thing's for sure, Terry's scope, pinger, and computers aren't
going to get much rest for the forseable future!
--
Paul Nicholson
--