Re: Primary Resonance

>> [ NOTE: Scott is currently on vacation. -- Chip ]
>Since Scotty seems to be on vacation, I thought I'd give a shot at
>answering your question. I'm running into a similar resonance condition.
>My system has 0.02 uF of tank capacitance and two 15 KV 60 MA
>transformers. It resonates at almost exactly 60 Hz, and shows a very
>sudden increase in primary current (jumping to 12 Amps) at about the 20%
>variac setting, and my gaps begin sporadically firing - sounds similar to 
>your system. However, I'm not seeing the "beat" frequency behavior that
>you are, so your system is probably resonating at a frequencly slightly
>higher or lower than 60 Hz. If we assume yours is slightly higher (say 62 
>Hz), you will see a lower "beat" frequency at 2 Hz (and may also see an
>upper beat frequency at 64 Hz). Even though the circuit is excited by
>60 Hz, you'll still see significant resonance effects since the "Q" of
>the neon-tank capacitor system is fairly low (typically 5-10), puttin 60 
>Hz well within the bandwidth of the tuned system.
>On my system, I actually see a significant drop in primary current as
>the effective "Q" of the resonant circuit declines when the gaps begin
>firing. With increased primary current, you may be seeing a slight
>reduction in the apparent series inductance of the transformers and
>variac as you have theorized. Since you are getting a higher than normal
>primary current for a given voltage input, you may also be starting to
>saturate one or more of the cores in your neons. This would reduce the 
>apparent inductance, and would cause an increase in the observed beat 
>But will I blow up my neons or caps???
>As long as you use a static gap adjusted to a break down voltage below
>that which would cause damage to the transformers or tank capacitors (I 
>use a maximum gap of 0.50") you will probably be OK for short-term, low 
>duty cycle operation. As you approach 60 Hz series resonance, the tank 
>capacitance partially cancels the current limiting effects of the 
>internal magnetic shunt inside the transformer. Under these conditions 
>you're supplying significantly higher current levels than the transformer 
>secondaries were designed for. While this makes for EXCELLENT coil 
>performance, due to more rapid recharging of the tank capacitor between 
>gap firings, it's also very hard on the transformers, and can also 
>overstress the capacitors. The heavier than normal secondary current 
>causes significant heating of the fine secondary windings. I've found 
>(the hard way..) that 15 KV 30 MA transformers seem to be significantly 
>more fragile than 60 MA ones when overstressed in this manner. 
>The SIMPLEST fix is to "overpower" your system by adding one more neon
>transformer to the bank. This will shift the new resonant frequency
>below the point where it will cause you a problem while still providing
>you with excellent coil performance. The next best solution is to reduce
>the size of the tank capacitance so as to raise the resonance point
>significantly above 60 Hz. However, this may reduce your output voltage
>and sparklength. With either fix, you'll notice that the variac setting
>will now need to be significantly higher before you the gaps begin to
>fire,  since you'll no longer be seeing the effects of 60 Hz resonant
>rise on the secondary. Your transformers will thank you for it by not
>smoking quite as quickly.... your capacitor(s) will have to fend for 
>[Note: I've also played with power factor correction capacitors connected
>across the primaries, and they DO reduce overall current draw. However,
>they do NOT fix a 60 Hz resonance condition].
>Safe coilin' to ya! :^)
>-- Bert --

Thanks for the input Bert. What you are describing is exactly what I was
saying. It is certainly annoying. From what I can gather it is something we
have to live with. I did try and add another neon and it behaved exactly
like you suggest. I have been struggling to understand why the beat freq
changed with increasing power and I think your saturation idea could well be
the reason. 

Thanks again. This has made things a lot clearer. Annoying as it is!