Re: Shrink wrapped secondaries

["Tesla list" <tesla-at-pupman-dot-com>]

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

Terry wrote:

> Without the shrink wrap, the ringdown looks like this:
> http://hot-streamer-dot-com/temp/02090602.gif
> http://hot-streamer-dot-com/temp/02090602.CSV

> With the shrink wrap it looks like this:
> http://hot-streamer-dot-com/temp/02090603.gif
> http://hot-streamer-dot-com/temp/02090603.CSV

Very nice recordings. The FTs look lovely,

http://www.abelian.demon.co.uk/tmp/02.ft.gif

for the unwrapped coil. The first six resonant modes show up very
nicely.  A few more are visible but not big enough to take numbers
from.  tcma reports 

Unwrapped 02090602.CSV:
 PK  FREQ kHz (Error +/-)    Q FACTOR (Error +/-)   LEVEL
  1  351.155 (0.01%,38Hz)     275.08 (0.17%, 0.5)  -0.3dB
  2  872.170 (0.01%,95Hz)     249.82 (1.66%, 4.1)  -13.9dB
  3 1263.176 (0.01%,137Hz)    190.67 (1.66%, 3.2)  -21.2dB
  4 1598.179 (0.01%,173Hz)    146.37 (5.58%, 8.2)  -25.5dB
  5 1904.784 (0.01%,207Hz)    117.08 (13.28%,15.5) -29.3dB
  6 2198.208 (0.01%,239Hz)     95.24 (26.56%,25.3) -28.3dB
 
Wrapped 02090603.CSV:
 PK  FREQ kHz (Error +/-)    Q FACTOR (Error +/-)   LEVEL
  1  351.160 (0.01%,40Hz)     275.08 (0.17%, 0.5)  -0.3dB
  2  874.085 (0.01%,98Hz)     251.05 (1.66%, 4.2)  -13.8dB
  3 1266.090 (0.01%,143Hz)    192.48 (2.35%, 4.5)  -21.2dB
  4 1601.364 (0.01%,180Hz)    147.36 (5.58%, 8.2)  -25.4dB
  5 1908.704 (0.01%,215Hz)    117.24 (13.28%,15.6) -28.7dB
  6 2204.296 (0.01%,248Hz)     96.10 (26.56%,25.5) -27.2dB

No change to f1 or q1.  The higher modes frequencies and q factors
have gone up a little - opposite to what we'd expect.  

I recognise this test coil, here is the comparison between your pings
and the tssp model.

     Measured   Modeled   Error
f1     351.2     356.6    +1.5%
f3     874.1     878.7    +0.5% 
f5    1266.1    1282.5    +1.3%
f7    1601.4    1626.2    +1.5%
f9    1908.7    1983.5    +3.9%
f11   2204.3    2303.7    +4.5%

The large error at f1 is probably because I don't know where your
coil is set up.  The increasing trend in error above f3 is due to the
model not taking account of the coil former material, and this causes
an error determining internal capacitance which increases with the
mode number.  This trend shows up very nicely and demonstrates the
benefit of pinging the coil to get a true picture of its mode spectrum.

Wonder why the frequencies and Q factors went up?  If ambient
temperature was involved, wouldn't all the Q factors change more or
less in proportion?  We would expect q1 at least to vary some 0.5%
per degree C.  Why does q3 go up nearly 1% and q1 stays put?

More curious still is the frequency change of the higher modes.  
Having gone the wrong way in the first place, why are the highest
modes affected most, trending towards a +0.3% shift in frequency.
If I see this pattern of error when modeling, it usually points to
an error in measuring coil diameter.  Has your shrink wrap compressed
the middle of the coil, reducing its diameter by about 0.5%?
The gain in frequency would go some way to accounting for the gain
in q.

One thing's for sure, you haven't upset the coil's resonant properties
by wrapping it, although you've managed to add a couple more mysteries
to the growing pile.
--
Paul Nicholson
--