Ariel: Works Again, then dies--I do nothing

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Dave Hartwick by way of Terry Fritz <twftesla-at-qwest-dot-net>" <ddhartwick-at-earthlink-dot-net>

Guys,
This is tough to explain. My instrumentation is not the best and has given
some difficult to interpret readings.

I am coinfident that the HV winding's DC resistance is about 560 ohms. LV
winding resistance is too low to measure. I forget the HV winding's
inductance--some henries--but I'm 89% convinced that the HV winding is not
the problem.

Consider the LV windings, henceforth referred to as the primary. As is
probably common, it consists of what I THINK to be 2 separate windings,
which allow for 120 or 240 V operation depending upon configuration. 4
copper straps issue from the core, terminating in the 3 terminals that
penetrate the case. The 2 inner straps are connected together to the center
terminal. This effectively results in a single center-tapped primary. The
schematic on the info plate is consistent with this assessment.

Here's where it gets mysterious: After the initial failure, I measure the
inductance across the
primary, and from center-tap to each outer terminals. From CT to left
terminal: 2.8 uH, from CT to right: 16.5 uH. This appeared to indicate a
problem, of course, so in somewhat of an act of desperation, I connected 120
VAC service and variac across the primary and applied some voltage.
The variac sparked and growled, similar to the behavior of the 240 variac on
240 V service right after the failuer, suggesting a serious short somewhere.
Then I tried it again and suddenly no variac sparking or growling--it
suddenly appeared to be behaving normally, and indeed it was. I was able to
draw an arc from the HV windings, and then operate my coil normally. I
reconnected 240 V service and variac and the coil fired normally. I
re-measured primary inductance. About 50 uH from CT to either side. The uH
magnitude is suspect--my Tenma LCR meter measures with a 120 and 1000 Hz
test signal and I'm not sure of how to interpret this, but I had symmetry,
consistent with a properly functioning center-tapped primary.

I then sought to duplicate a primary strike, but this time with the safety
gap 100% confirmed to operating properly, which it was (The center electrode
was not grounded during failure--failed aligator clip spring!) I set a
coffee can on the outer turn of my coils primary, and ellicited several
primary strikes which trigered the safety gap and were drained to main RF
ground. Then a final primary strike, and Ariel quit again.

I once again measured inductance, and got the assymmetrical readings, 2.8
and 16.5 uH.

I removed the lid and pulled the 4 copper primary lead straps from the
terminals. Measuring for continuity this time, I see there is continuity
between the now disconnected 2 primary windings, which I assume should not
be, but inductance measurements of either side show about 7 uH. I did not
measure for continuity during the short interim in which the xfmr decided to
work once again. I should have.

So what we have is a transfromer that is intermittent--from 100% operation
to nothing but what appears to be a dead short, probably in the primary. I
really have no choice but to pull the core at this point. Any guesses? This
is a mystery.

Thanks for enduring the length of this.
Dave H