You might also wish to try a "parallel plate transmitter waveguide" like I
used in the early 1960s. It consisted of a wide strip of acrylic sheet,
0.250" thick, and, then on each side I affixed a wide flat copper ribbon,
thus forming a capacitive "transmission line" allowing me to transfer
power
to the coil base. It had an extremely low inductance (less than 2 uH) at
distances up to 30 feet. Over a 10 ft. distance the inductance would be
very low. One drawback was an occassional flashover from sec HV terminal
to
the transmission line. We were using very robust capacitors in our
oscillator so no damage was ever sustained by the transmission line
strikes. My acrylic was 8 inches wide and each copper strip was 5 inches
wide to allow a very long creepage distance. All sharp edges on the
copper
strip were deburred and the edges coated with G.E. Red Glyptal.
This was later abandoned when "common sense" dictated I place all the
components directly under the coil. I began using this design around 1971
and have used it ever since. It ended the flashover problems except to
the
HV feed lines. Metal raingutter( lay out inverted) is cheap and prevents
this problem in large coils.
Dr. Resonance
2009/8/2 Peter Terren <pterren@xxxxxxxxxxxx>
Thanks for all the suggestions. I did switch to a coax setup for primary
lead wires with heavy wire inside about 50cm copper pipe to each primary
and
I think that it improved output over the parallel conductors.
Interesting about the coupling because using a single secondary, I could
induce racing arcs before if I tried to max coupling.. Not as twins,
however, although the 6-7 turns required probably puts the coupling range
back to "normal". I also have plastic wrap on the coils as well.
My application does not permit separate electronics at present and the
leads need to be almost 80 cm.
I am aware of the V-twin at Palais. I already had the boards and
components for a standard SISG though but a triggered SISG sounds
attractive, particularly if I could use some of my IGBT bricks. The low
turns and high current would require the electronics to be separate and
local which is not suitable here.
So at present I have 80cm sparks between 50cm secondaries. Certainly
nothing to write home about but is probably adequate. I am awaiting
better
SIDACs as half were 200V rated and plastic packed so starts firing at
about
180V on the variac (out of expected 250V) Had to source them from China
as
the usual suppliers were out of them. Hopefully that will take sparks up
to
100cm.
I did flame a MOT today with this arrangement so some voltages are
appearing when they shouldn't. Plenty more where that one came from.
Peter www.tesladownunder.com
----- Original Message ----- From: "Bert Hickman"
<bert.hickman@xxxxxxxxxx
>
Hi Peter,
The stray inductance of your primary circuit is reducing the effective
coupling of the twin system. There are several possible fixes, You've
already tried one (reducing primary tank capacitance and reducing stray
inductance).
You could also try increasing the coupling by elevating the primaries
(only when running as twins). Stray inductance is reduced by reducing
the diameter of the current loop in the primary circuit(s) by bringing
supply and return current paths closer together so that their combined
magnetic fields tend to (mostly) cancel. Some possible options include
running them closely side by side (like lamp cord or as a twisted pair),
making a sandwich HV stripline structure with copper strips separated by
a HV dielectric. One of the best configurations would be to make the
equivalent of a high voltage/high current hardline coax (as you
suggested), using a solid outer copper pipe and heavy gauge silicone HV
wire inside or and inner conductor of 1/4" tubing with insulating
spacers. This configuration should minimize corona losses versus
stripline.
You're not looking so much for impedance matching, per se, but canceling
the stray inductance that messes with your tuning and coupling.
Bert
--
***************************************************
We specialize in UNIQUE items! Coins shrunk by huge
magnetic fields, Lichtenberg Figures (our "Captured
Lightning") and out of print technical Books. Visit
Stoneridge Engineering at http://www.teslamania.com
***************************************************
Peter Terren wrote:
I am working on a small twin coil system with a cylinder primary and
small spherical topload. Using a single coil only, run by a single MOT
SISG, it puts out a 60 cm spark for a 50 cm secondary winding.
I am finding results a bit limited when run as a twin system due to the
length of the primary leads. With a series primary arrangement I
should
tune to 3 turns on the primary but with loose leads this drops to about
1.5 turns which is not efficient and spark length is less than 30cm
despite double the power.
With parallel primary windings, I should tune to 6 turns but it is
reduced to 3 turns with the lead in wire inductance. It seems like I am
wasting half of my inductance.
Best performance of 90cm sparks with 2 MOT SISG is with a reduced tank
capacitance to allow greater inductance in the primary using 8 turns in
parallel connected primaries with lead in wires taped together sort of
transmission line like.
My question: How can I minimise the effect of primary lead inductance.
Should I use two close parallel conductors (sort of transmission line
like) or would a coax arrangement be better? I can make up a coax
system with copper pipe and plastic tubing covering 1/4 inch tubing. I
don't really understand the concept of impedance matching as applied to
Tesla coil primaries though or whether it is even likely to be
relevant.
Any thoughts?
Peter www.tesladownunder.com