Capabilities of Small Thin Wire Secondaries (fwd)

---------- Forwarded message ----------
Date: Mon, 27 Jul 1998 17:57:23 EDT
From: FutureT-at-aol-dot-com
To: tesla-at-pupman-dot-com
Subject: Capabilities of Small Thin Wire Secondaries


The purpose of these tests was to explore the capabilities and
limitations of small thin wire secondary coils. 

Initial specs:   Pri, # 12 stranded pvc ins. wire close wound flat
pancake, 3.5" inside dia.  tapped at 18 to 21 turns as needed, about
87uH.  Sec, 3" by 9" styrene form tight wound with 1200 turns # 34 
isomid enamel wire, 30mH.  Toroids, 1.5" by 4.5" placed on top of
sec, with a 3" by 12.5" alum. dryer duct toroid on top of the small
toroid, with metal spacer in between.  Static gap is (24) 1/2" by 2"
copper pipes glued to a board, spaced 8.5 mils similar to Terry F's 
design.  Powered by one or two 9kV 30ma NST's.  Cap is .007uF 
polypropylene.  No safety gaps, chokes, resistors, etc.   

Test 1:  For this test, I used an old static gap having 4 series gaps,
each spaced about 1.2mm.  This gap performed poorly and quenched
on the second notch at k = .1 (sec is 1.25" above the pri.).  Spark
length was 21".  Racing sparks appeared on sec at tighter k values. 
One NST was used for this test.  Sparks are blue and do not appear
purplish despite the thin wire.

Test 2:  Cap was increased to .014uF and coil re-tuned, but racing
sparks made it impossible to continue.  With sec raised 2" above
primary, output sparks were weak.  .014uF may have been a little
too much capacitance for the NST also.

Test 3:  Back to .007uF using the new Terry F.- type gap, (used
14 gaps).  This gap quenched much much better than the previous
4 section gap, and gave 1st notch quenching at all times even with
tightest coupling at k = .2 (sec lowered all the way).  Sparks are
still 21" long, but now multiple streamers are formed.  Streamers
occasionally strike the primary.  A few racing sparks appeared
after awhile, but a household fan blowing on the gaps helps slightly.
Eventually the gaps heat and racing sparks appear, although the
quench on the scope could be seen to remain at 1st notch.

Test 4:  Installed 4" by 17" toroid on top of small 1.5" by 4.5" toroid.
Now only one streamer was formed, but a "corona column" formed
between the primary and the toroid at one point.  I had to raise the
large toroid using a metal spacer to prevent this.  23" sparks were
obtained.  Racing sparks form on sec after the gaps heat, but scope
still showed 1st notch quenching.

Test 5:  Re-installed the 3" by 12.5" toroid, tuned at 19.24 turns of
pri., streamers reach 22.5".  K = .18  (had to loosen due to continuing
racing spark problems).  At low power, the gap quenched at the 2nd
notch, gave 1st notch quench at higher powers, due to better loading,
and tuning at the higher powers.  I noticed some corona flaring 
between the pri and sec coils (only 3/8" spacing).  The gaps fired
twice per ac half cycle in this and the above tests. (Did not check 
the firing rate using the .014uF cap, but was probably 120 or 60 BPS).

Test 6:  Installed (2) NST's instead of one, but kept the gaps the 
same, which gives a higher gap firing rate.  Sparks hit 25".

Test 7:  Installed the 4" by 17" toroid.  Gave about 26" sparks,
very strong and dramatic looking, but sparks occasionally lept
from toroid to primary.  Eventually the secondary broke down turn-to-
turn at one point and burned through the wire.

Test 8: Removed a few turns of burnt wire, and resoldered the ends
thereby repairing the secondary.  I also removed the inside turn of
the primary to give a little more clearance, and retuned to compensate.
Installed a metal ball in place of the lower small toroid and ran the TC.
Sparks reached 24.5", but racing sparks threatened to destroy the sec.
I tried some different toroid heights, etc, but problems contined.

Perhaps some tinkering with the toroid positions, etc, could help
the situation, but I suspect the secondary may be near it's limit when
producing around 25" sparks.  The TC reliably produced 21"
sparks without breakdown problems.  It may be possible to narrow
the total gap spacing, and use a higher current transformer, to raise
the break rate, and thereby increase the maximum spark length with
this secondary, but I didn't try it.

Next, I installed my 3" by 12" secondary wound with #28 formvar
wire for comparison.  Although only the secondary was changed,
and the coil retuned, other parameters unavoidedly changed too.
This new secondary has about 790 turns, 11.2mH, height is 
different, and Fo is different.  Also the surge impedance of the
primary is lower with this set up.

Test 9:  Installed the 3" by 12.5" toroid. and tuned at turn # 10,
gave 21" sparks which emited from a few places on the toroid.
Sec is lowered all the way for k = .16 or so.  (one NST).

Test 10:  Installed two NST's,  gave 22.5" spark.  Had to raise the
secondary slightly to give k = .145, for best results.  No racing sparks,
and spark does not hit the primary.

Test 11:  Installed the 4" by 17" toroid over the 1.5" by 4.5" toroid,
with flattish spacer in between.  Obtained 25" to 26", and no racing
sparks appeared (tuned at 12.25 turns).

Conclusions:  The tuning seemed less critical with this secondary,
and racing sparks and other types of breakdown did not occur, even
when the gaps heated up, and despite the fewer primary turns in use.
Quenching occured consistently at the first notch.  Spark lengths
seemed the same for both secondaries.

Next, I installed a 12kV, 30ma NST, but the extra voltage
caused uncontrollable racing sparks.  All 24 gaps were used.  

Since higher voltages or a larger capacitor both caused racing sparks,
in these tests, using these secondaries, the only way to increase the 
spark length was to increase the break rate by using a narrow total 
gap spacing, and a higher current transformer.  I could have installed
a larger secondary which would permit a higher voltage/larger cap to
be used, but I did not want to do this, since it would be outside the
planned scope of these tests.

Test 12:  Out with the NST's!!  I installed a 14.4kV, 1.5kVA potential
transformer and an adjustable inductive ballast.  16 gaps were
used to keep the firing voltage low.  Still using the 4" by 17" toroid,
the sparks easily hit 32" at 800 watts (1000VA).  The sparks streamed
straight outwards and did not hit the primary.  After running awhile the
gaps heated up and some racing sparks occured on the secondary,
but the scope still showed 1st notch quenching.  After awhile, the
gaps fouled which reduced the spark to 16" or so, and the gaps had 
to be cleaned to restore proper operation.  In one run, the sparks
reached 34.5", but the power level may have been higher.

Conclusions:  A short secondary can produce a strong spark, but
the voltage and "bang size" has to be kept low, and a high break
rate used, to grow the sparks.  This method may less "efficient"
however than using a somewhat larger secondary along with a
higher voltage and/or a larger capacitor, and a lower break rate.
As an example, my 4.25" by 23" secondary (# 28 wire), gives 46"
streamers, using a .014uF cap, with a 5" by 20" toroid, at 60 breaks
per second, and uses 600 watts. 

Next, I installed my 4.25" by 23" secondary (#28 wire), and a
larger primary (# 12 stranded, pvc ins., ~8.5" inside dia., flattish
15 degree inverted cone).  This primary was too large for the sec, so
the coupling was limited to k = .1 (secondary lowered to 2" below the

Test 13:  The gaps, etc, were left unchanged from test 12, primary
was tuned at 19 turns.  The 4" by 17" toroid was used.  The sparks
seemed to be slightly stronger than with the 3" by 12" secondary,
at the same power level of 800 watts, but it was a little hard to tell
because the higher primary surge impedance required re-adjustment
of the ballast inductance for steady operation, but some unsteadyness
remained.  This sort of unsteadyness was not seen with previous
sync-gap operation, so I suspect that static gapped systems with
high surge impedance primaries demand more critical ballasting,
and possibly some resistive damping (I didn't try it here yet).  It is
also possible that I needed a greater inductance range for the 
ballast, but I didn't check this out yet. 

Test 14:  The input power was turned up to about 1200 watts, and
the sparks easily reached 37".  No racing sparks or breakdowns

Test 15:  The number of static gaps was increased to 20, and then
to 24, but the coil ran unsteadily.

Conclusions:  This last group of tests suggests that the large
secondary has improved the performance of the TC slightly, but many
parameters have changed, (k, primary surge impedance, etc.).  It
is clear that the performance is poorer than with my old design with
the sync-gap which gave 42" sparks at 620 watts.  The only
difference in the sync-gap set-up was the use of the sync-gap
instead of the static gap (and the lower break-rate that resulted
from using the sync gap).

Relatively small secondaries are useful in the design of compact,
but relatively powerful TC's, for portable, demo-type use, or where
space is limited.

John Freau