Strange safety gap operation (LONG POST)

Hi guys,

I am currently spending time pushing my 200BPS sync rotary coil to
progressively higher power levels.  Since the break rate is fixed,
I am increasing the power by increasing the HV supply voltage. This
is being doing gradually using an HV scope to set the rotary
phase correctly at each new supply voltage setting.  While doing
this I found something which I think is strange but others may have
experienced already.

ACROSS THE ROTARY is a safety gap consisting of two parallel copper
tubes.  I set the safety gap by connecting the HV scope accross the
safety gap with a gap of initially 1mm, (40thou,)  and I manually
rotated the rotary so the electrodes were not lined up,  and turned
on the HV.  The safety gap roared to life,  firing many times per
cycle but only at a few kv.  Over a series of tests I opened the
gap progressively until I got it to fire very smoothly, clamping at
my chosen point of 12Kv.

After the safety gap was set as explained I thought no more about
it and ran the sync rotary many times with the scope still connected
showing the cap charging to exactly 10.5kv at 200 times per second.
The TC ran very smoothly and the safety gap has never fired except
for when I once forgot to start the rotary first :-<
(Thank god I fitted a safety gap.)

After demonstrating the system at the Teslathon making a 30" spark
(and a characteristic whining sound,) I did a series of tests
incrementing the peak capacitor voltage by 1Kv each time,  until
18 kv was reached.  With the capacitor charging to 18kv at 200 times
per second the operation was smooth and still the 6mm safety gap
never fired once !!!

If I set the safety gap accurately to 12kv, why had it not began to
"cut-in" when the voltage accross the rotary was now 18kv just
before each firing ?

This seemed very odd to me so I boldly repeated my originaly test
by misaligning the rotary electrodes and hitting the HV switch
with rotary stationary.  This time with the much higher voltage
supply.  The safety gap fired as expected and the voltage was
firmly clamped at 12.2 kv despite the higher voltage supply !

Will the safety gap not fire if the overvoltage is "regular" as
in the case of my rotary gap ?

I originally thought that the way in which I set my safety gap
threshold was a good method because I could actually check the
clamping voltage as I set it,  but now I suspect that a single gap
has a different breakdown voltage when not firing to when firing
regularly.  Can anyone else confirm or repeat this test ?

I have given much thought to this,  and it could be due to the
instant change in temperature from not firing to firing.  This
is however very important because it implies that the threshold
voltage will be far higher when "cold" and will drop when warm.
This means that the first firing of the gap won't occur until
the voltage is VERY HIGH and then following firings will
occur at a lower voltage when the gap is "warm".  (Not very useful
for my safety gap !)

For static gap users this implies that there will be a
significant overvoltage transient when first firing up a coil.
(Maybe it is better to switch on the supply sharply rather than
winding up the supply slowly on a variac.  Get the overvoltage over
quicker by starting the static gap firing quickly ?)

I have not been able to confirm this because I don't have a
storage scope,  but I can recall making three observations when
previously using a RQ style static main gap:-

1. The variac would have to be turned up high to start the static
   gap firing,   however it could be turned down several percent
   before the gap would stop firing.

2. The only 2 capacitors I have lost both failed instantly on
   power up before the gap fired. (OK the gap was quite wide.)

3. If forced air-cooling was switched off when using an RQ-gap
   the gap temperature would rise and spark length would dive.
   I found this was due to more rapid firing of the gap at a lower
   voltage,  giving less energy overall.

All of these points (particularly the first)  suggest that the
static gap threshold is quite variable and the way in which I have
set my safety gap is not good.  On the other hand 6mm sounds quite
small for 18kv to me,  I have not blown anything yet,  and the
rotary runs very smoothly.   Comments needed,  please.

As a final note two people told me that point three above is due to
poor quenching in the overheated gap.  I thought "quenching" refered
to the turn-off ability of the spark gap.  This problem sounds more
like lowering of the breakdown voltage due to overheating ?

I hope I have not rambled too much,  but I feel that understanding
the properties of the static gap is very important to coilers
whether we use one as our main gap or as a safety gap.

			- Richie

			- Somewhat puzzled in sunny Newcastle