Re: cap firing voltage scope measurements question

Hi John,  and all,

I think this behaviour you mention is quite normal,  especially if you are
using a stiff power supply,  small capacitor,  and a low break rate.
The capacitor has to charge to a high value to process the available power
if the rotary phase is set correctly.

I did some simulations and guessed that your transformer would give around
130mA s/c at 10.6kv,  and the cap was around 15nF.  The rotary phase was
probably near optimum.  Did the scope waveform show a smaller voltage
swing in the same direction just after each bang ?  If so,  this indicates
the reason for the resonant voltage rise.

After the gap fires, the capacitor charges considerably on the down side
of the mains waveform but does not fire.  The cap voltage then swings back
to a higher voltage of opposite polarity when the opposite supply cycle
begins.  This involves energy storage in the ballast inductance.

More comments below:-

On Wed, 16 Jun 1999, Tesla List wrote:

> Original Poster: FutureT-at-aol-dot-com 
>  All,
>  I'm seeing some strange results in my TC:
>  Here are some cap voltage measurements under different conditions:
>  I set my input variac at 60, and the scope indicated the output voltage
>  of xfrmer was 15.2kVpeak, without the TC connected.
>  At the same variac setting, but with TC connected, but with the 
>  spark gap removed, the cap voltage reads 22.3kVpeak.  I suppose
>  I'm getting some resonant charging here with this particular ballast
>  setting.

There is only a small resonant rise so the capacitor value must be some
way from mains resonance.  My guess is much smaller capacitance.

>  At the same variac and ballast setting, but with the gap firing 
>  (120bps), and the TC producing sparks, the cap voltage reads
>  30.4kV.  This means I'm obtaining a 2X voltage increase during 
>  ~1/2 cycle, I thought this degree of resonant build up was possible
>  only when using DC charging?

I take it that the 30.4kv is the peak value just before the bang ?
This voltage rise is to be expected due to the effect of the gap firing.
I have only once experimented with DC charging at low power,  and never
experienced any resonant voltage rise effects.  Would anyone else care to
comment on DC systems ?

>  Does this seem normal?  I would have expected the voltage to be
>  lower with the gap firing since there's less time for resonant voltage
>  build-up.  In there a gap in my understanding?  Or is my scope
>  reading wrong?  The 30.4kV makes sense from a Joule perspective,
>  since it comes out to 840 cap watts or so, and the meter reads 1000
>  watts, which seems about right. 

Firstly,  trust the scope !  Try verifying the cal on a known HV supply or
on the mains supply if you are still suspicious.

As you say the figures agree,  and I have seen this occur in simulations.
This phenomenon may also explain why safety gaps persistently fire at
certain settings event when the rotary is set correctly.

>  Mystery, or insanity?

There is still much to be learned about the complex interactions of
resonant charging systems,  but a good way to estimate the peak voltage
with a correct rotary setting is to take the measured VA of the supply and
divide it by the break rate,  etc.. to calculate the necessary capacitor
voltage and bang energy.

						- Richie,

						- In sunny Newcastle