Tube coil makes sparks

Fellow Coilers,
   Just thought I'd fire one off to report that, after my tortured parts
seeking exercise, the 833A coil is running and producing vigorous 20"
sparks. Fr = 230 khz.  But this is with 3 paralled tubes. Lots of
fiddling to do, obviously.
   John Freau has noted that high power tube coils produce a discharge
whose character is "angrier" (I think that's the word) than that of low
power coils. I can confirm this in abundance. The lower half of the
sparks are thick, white, and very satifying. The upper halves are
thinner, and a vivid sly-blue. The audio output is quite impressive-- a
robust 60 cycle note accented with artifacts of the discharge that can
be "poppy" and/or spitty. It  would be interesting to power the coil
with DC and modulate the output with   a 40 cycle tone. A second power
transformer could be used as a modulation transformer, I think. ( I know
hams have used TV power transformers as mod xfmrs in AM transmitters
modified for wide-band response.)
   There has been some debate over the usefulness of the MMC capacitor
arrangement. I am currently using 6-- 0.018 uF, 1600Vdc, Panasonic
metallized polyprop caps in series for the tank capacitor (result =
0.003 uF). They were  mildly warm after a 5 minute run at 4000 Vac and
probably 4+ kW.  I was hoping they would run stone cold. A permanent
setup could incorporate 24 caps in a series parallel arrangement, for
example. Despite the modest heating, I must tentatively conclude that
they are satisfactory for high power tube coil duty.

   I have in my possession a total of 4-- 833As. It turns out that one
of these produces a beautiful blue glow consistent with John F's
description of a current sucking corpse that is distinctly unusable. Of
the 3 tubes remaining, 2 have graphite plates. To say that they are more
stout than those with conventional plates is an understatement---where
the conventional plates are bright orange, the graphites are barely
glowing. I may need to bias the tubes separately.
   I'm using a 100 watt, 10K variable Ohmite resistor for grid-bias
duty. $ 25.00 from Digikey. I almost bought the 225 watt version. Wish I
had. The 100 watters can get quite hot, especially when  less than half
of their resistance is utilized. In retrospect, it may have been better
to obtain several fixed value units so that the full heat dissipation
capacity of the single resistor may be mobilized.
  So there is my report. Much work to do. The cancerous desire for more
power was felt very early on. I know that 3 tubes will produce much more
than 20" sparks, but a gentlemen at Elex electronic supplier mentioned
the availability of 10,000 watt industrial oscillator tubes. ...
Dave Hartwick