Strange transformer problem

I've run into a situation I've never seen before and hope some of
you magnetic gurus can help me figure this one out.

I'm testing a new setup involving using a medium sized variac as a
variable choke to control the current to a 25kva pig.

the following components are in series across a 240 volt line:

* 25 amp, 240 volt variac connected between one winding end and the
wiper (John Freau- the one I bought from you)
* Fixed choke consisting of 175 turns of 0.2" dia magnet wire around
a bundle of 7 lbs of 18" long mild steel gas welding rods (about an
inch in diameter.)
* 25KVA, 14.4kv/240 volt pole pig.

In testing this setup with the pig secondary OPEN, things progress
smoothly until the variac reaches about mid-range.  Then it starts
stuttering.  That is, an irregular growling sound.  An ammeter in
the primary shows spiked current draw with each growl.  sounds like
a gap when it is just breaking down and the spark is still blue and
irregular.  Advancing the variac a bit more makes the growl become
steady and the primary amp draw is about 1.5.  Advancing toward the
stop gradually reduces the growl AND the primary current so that
when the variac is at the stop and therefore out of the circuit, the
magnetizing current is less than a quarter of an amp.  

Here's the really weird thing.  As I turn the variac back down, when
it reaches approximately mid-scale, the growling returns, the
magnetizing current goes up AND the voltage applied to the pig rises
from below 240 volts to about 275 volts.  If the variac is gradually
turned all the way down, this behavior continues almost without
change so that when the variac is all the way against the lower
stop, the growling is again steady and the voltages are as follows:

* line - 240 v
* across the pig - 275 volts
* across the fixed inductor - 9.5 volts
* across the variac - 360 volts!

Measured with a Fluke 88 and checked with a Beckman DVM, both in
current calibration.

The primary current is again about 1 amp.  Once this condition
establishes itself, jumpering out the fixed inductor makes no
difference.  remember the pig secondary is OPEN and there is no
capacitor anywhere in the circuit other than distributed

So what's going on here?  The way the variac starts to growl sounds
like a breakdown but it meggers good.  The pig's been in service on
my neon bench for a couple of years so I know it's good.  when the
variac is taken out of the circuit and connected conventionally, the
voltage rises smoothly as the rotor is turned.

This 'un has me baffaloed! 


John De Armond
Neon John's Custom Neon
Cleveland, TN
"Bendin' Glass 'n Passin' Gas"