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

1st notch tests, continued

Subject:  1st notch tests, continued
  Date:   Wed, 16 Apr 1997 03:39:40 -0400 (EDT)
  From:   FutureT-at-aol-dot-com
    To:   tesla-at-pupman-dot-com

All,

I finally got around to metering my input current on the 12kV, 30ma
set-up.
 Meter shows 3.25A at 240V, or 780VA.  I use a 240V variac on a 240 volt
line
and turn the variac up about half way for 130V into the transformer. 
This
neon transformer is a common, run of the mill, Jefferson or Magnetek. 
This
one has never been blown out or rebuilt.  I use a 60uF PFC cap to help
reduce
the VA draw.  These results show that a neon trans can draw a lot more
than
its rated current due to resonant charging.  When comparing the
efficiency of
different neon TCs, it necessary to state the measured input VA draw.  

For the first coil test, I lowered the secondary to 1/4" below the
primary
for a k of about .19.  No racing sparks appeared, quench may have been
on the
first notch, but had a lumpy look, as if it was between magic values,
but it
may have been alternating between good and poor quenching, or may have
been
3rd notch -- not sure.   Sparks are shorter due to multiple streamers.

Next, installed a 5" x 20" toroid on top of the 4" x 20" toroid, gives
one
streamer now, and spark is longer, but still does not hit ground wire
which
is 38" away.  Quenches on 2nd or 3rd notch.  Raised sec to 2 1/2" above
pri,
(k ~ .11), spark is stronger and hits ground wire, quenches on 2nd
notch, and
gives one streamer.  If k is loosened, spark weakens, but still quenches
at
2nd notch.  If k is tightened, spark weakens, and gap quenches on 3rd
notch.

Installed the 5" x 20" toroid ONLY.  With sec 2 1/2" above pri, gives
good
spark, one streamer, and quenches at 1st notch.  Lowered sec to 2" above
pri,
spark increased to 39 1/2" but quench is starting to look lumpy and may
be
tending toward 2nd notch.  Lowered sec to 1 3/4" above pri, quench may
be
tending toward 2nd notch, but spark is stronger and hits 40 1/2" 
Lowered sec
to 1 1/2" above pri, --similar to above, but may be 2nd notch quenching.
 Lowered sec to 1 1/4" above pri, gives 41" on a regular basis, and hit
42
1/2" one time with a strong strike.  This is the best results I've
obtained
so far using a 12kV, 30ma neon trans.   Lowered the sec to 5/8" above
pri,
but spark is weaker.  Lowered sec to 1/2" then 1/4" above pri.  Sparks
weaken, and quench moves to 3rd and 4th notch respectively.

Conclusion:  It seems that now that the gap quenching has been improved
(see
previous posting), the system is able to benefit from the larger toroid
and
the toroid's ability to create only one streamer at tighter k values (up
to a
point),  k probably = .13 or so at sec = 1 1/4" above pri.  In any case,
the
better quenching seems to allow the system to benefit from the use of a
larger toroid.  It also seems that there is very little difference in
spark
length whether gap is quenched at 1st or 2nd notch, but 3rd notch quench
may
start to reduce the spark length.  These results tend to agree with the
higher power work I did at 2680VA, but it is not known if the general
thread
of these findings will apply universally to all systems.  Another way to
summarize these results is:  For best spark, use a toroid size which
will
still give 1 streamer, at the tightest k factor, which will permit 1st
or 2nd
notch quenching.  The limiting factor here seems to be the quenching
capability of the gap.  With a better gap, it may be possible to tighten
the
coupling, and use a larger toroid, for longer sparks.  

In a final last test, I removed all the toroids, and ran the system with
only
the 4 sharp metal "prongs" (sticking outwards at the top of the sec),
that I
use as a toroid holder.
Without a toroid, the system always quenched easily at the 1st notch, at
any
k value up to the max of .19 (?) that I can physically attain, (I did
not
check to see if removing the toroids alters the k factor).  Also, slight
mistuning did not prevent 1st notch quenching.  The spark increased from
18"
to 24" as the sec was lowered to the tightest coupling position.  This
seems
to imply that smaller toroids result in easier quenching than larger
ones.
 Some of the other findings above or in previous postings also support
this
view.
  
In today's tests, I really didn't see much or any benefit from the use
of
magic k values; changes in quench quality, streamer length, and
formation of
one streamer, with changes in k seemed quite linear.

Towards optimal coiling, comments welcomed,

John Freau