Poulsen first light

["Michael H Nolley" <nolleym-at-willamette.edu> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)]

    Greetings!
    Today I tested my Poulsen Arc TC for the first time.  The results
weren't spectacular, but I still have some kinks to work out.  Here are the
specs.

Supply:  2000vac 1A Mot (very large), rectified and filtered through a 41 mf
2kvdc capacitor, and a 100kv L diode filter for transients.

Gap:  Poulsen Arc gap, consisting of a 1 inch diameter water cooled copper
electrode, and a 3/4 inch diameter graphite electrode.
    At right angles to the gap are two modified Mot cores ("E" cores) which
contain 2 Mot secondaries each, for a total of about 8000 turns.
    The center leg and one of the side legs of both E cores are extended
into the chamber by laminated pole pieces ("I"s).  The center legs are the
operating pole pieces, and consist of short "I"s.  They terminate with nomex
insulated NIB magnets just short of the gap.  The side legs, which are on
the bottom, are connected with two complete Mot "I"s, which run directly
under the gap.  These complete the magnetic circuit.  The second side leg of
both Mots is unused.
   The whole apparatus is lashed together with PVC insulated copper wire,
since I couldn't find a way to weld the Mot cores to the pole pieces without
losing the ability to change inductors, and without melting the chamber (an
HDPE bucket).
    Eventually, when low power tests are over and I have found the proper
metal chamber I may weld them.
    A small alchohol lamp sits on the bottom pole piece and burns in between
the gap, supplying the hydrogen atmosphere.

Cap:  2 1600V .01mf Panasonic caps in series.

Primary:  8" diameter helical primary, using 16ga.  PVC insulated wire.  21
turns max.

Secondary:  6" diameter 26" long secondary, resonant frequency around 330khz
unterminated.

    I started out by limiting the supply power with a 10k resistor--this is
the only way I could run it without tripping the breaker.  The gap wouldn't
ignite at all--I would move the graphite electrode into contact with the
copper one, and there'd be a "chirp" and then nothing.  Depending upon the
capacitor size I'd get anywhere between 1 inch and 2 inches during these
moments.  I think the culprit is the inductors and possibly the
resistors--not enought current is getting to the gap to sustain the arc.
The first thing I'll try is limiting the primary side of the Mot instead of
the 10k resistor.  Then I'll probably cut each inductor down to 1 Mot
secondary--and if this doesn't work I'll try primaries.  I'm afraid this may
be too little inductance, however.  Another possible route is higher
voltage, which I may try even before this.  I have five more 2kv 1A Mots
which I can configure for 8kv 1A (without doublers).  In the next day or so
I'll be doing higher voltage runs, and I'll keep the list updated.

            --Mike