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

Archimedes Spiral Primaries-Spiral Magnifier Drivers


----------
From:  Dave Sharpe [SMTP:sccr4us-at-erols-dot-com]
Sent:  Monday, August 24, 1998 2:39 PM
To:  Chip Atkinson
Subject:  Archimedes Spiral Primaries-Spiral Magnifier Drivers

Bert H., Chip, All

One additional nugget of information has appeared involving
"spiral coils".  At work, a consulting company issued a report
of using spiral coils in a proprietary process application.  One 
of the "bottom line" comments was that the optimum geometry for
maximum Q for a flat or saucer-shaped spiral coil was when the ID
was 1/5 the OD.

This finding was reported in the following references...

-M. Adler, "A Field-Theoretical Approach to Magnetic Induction
 Heating of Thin Circular Plates," IEEE Transactions on Magnetics,
 vol. MAG-10, no. 4, December 1974, pp. 1118-1125.

-R. Chaddock, "The Application of Lumped Element Techniques to High
 Frequency Hybrid Integrated Circuits," Conference on Hybrid
 Microelectronics (1973, University of Kent), pp. 209-220.

Another un-referenced comment was that the coil have as wide a 
conductor path as possible for maximum Q.  Flat copper/aluminum
banding would be excellent in this application, with the only negative
being possible corona at the sharp edges. Banded stock would allow
more turns within the max Q geometry limits mentioned earlier,
possibly allowing a wider useful tuning range. Many of us have used
copper tube stock (I have 4 primaries built as such) and have gotten
excellent performance using round cross section material.

There has been an ongoing thread concerning spiral coils used as a
primary/secondary driver combination for small to medium sized
magnifiers. I built a small 450VA NST magnifier using a flat primary
and parabolic cross section secondary that could be adjusted axially
to adjust coupling.  I did achieve very high couplings (>0.55) without
magnetic flux concentrators.  Another interesting outcome was
the coupling varied linearly in three axes: That is height,
active primary turns (tap [tune] point), and measured coupling.
This was quite unexpected but did help explain why very early radio
used pancake coils for tank and antenna coupling coils, k could be
easily
adjusted based on tap position and distance between coils. The coil did
function as a magnifier, and sparks approaching 18" were observed.
This unit was shown operating on one of R. Hull's TCBOR Video Report
Tapes.
Other salient points I remember (this coil did a famous crash and burn
BURRAPT)...

Transformer was 15kV, 30mA NST
Capacitors ~0.007uF (2) paper/poil film "homebrew" under oil
   Tesla Equidrive topolgy, with gap across transformer.
SG 10 point air blast static gap, total gap ~0.15 - 0.20"

The negative was very high potential gradients between both coils would
lead to sudden and occasionally violent flashovers between primary and
driver secondary.  This was occurring through 2 layers of 1/4" thick
switchgear grade fiberglas and one layer (0.125 thick) switchgear grade
ABS material, and 1/2-3/4" of air. Unless oil immersed, this would be a
very problematic area IMO. I would expect similar problems with early
high
powered SG wireless transmitters in shipboard applications.  High
humidity,
corrosion, salt spray/contamination sounds like a loud BURAPT to me
somewhere
up the road...

An alternative design, shared with Dr. Mark Rzeszotarski, was use
a regular helical coil primary and a flat pancake secondary. The OD of
the
pancake secondary is grounded and the helical primary is powered as
customary.
The centerline of the pancake secondary was centered with the centerline
of
the height of the primary helical coil.  Mark indicated good results
with
such a system (Mark, jump in!)  The voltage gradient would then be
across
the spiral secondary, possibly leading to racing sparks radially across
the spiral secondary.  Again, oil immersion would help control these
voltage
gradients...

FWIW.

Regards

DAVE SHARPE, TCBOR