[Prev][Next][Index][Thread]

Re: Capacitors as current limiters



Hi Jim, all,

In theory capacitive ballasting would work in a steady state AC
supply, however there are a few reasons why it does not work in
this type of disruptive TC system:

1. Resonance with transformer primary inductance.  The reactance of
   the ballast capacitor is of the same order (but opposite phase)
   to the reactance of the transformer primary.  They partially
   cancel each other. The result is a high current flow,  and a
   large voltage rise across both components when unloaded.  I
   actually tried this a long time ago, without thinking first.
   The transformer growled loudly, then there was a huge bang as
   the PFC cap vented.

2. Normally a TC system appears as a slightly inductive load to
   the AC mains supply. This is because the tank capacitor partially
   cancels out the ballast inductance,  so the power factor is quite
   good.  Also, it can be further improved by adding PFC capacitors
   across the line to cancel the last bit of remaining inductance.

   In the case of capacitive ballasting,  the system appears as a
   highly capacitive load to the supply.  This gives very poor power
   factor,  (like less than 0.1) which cannot be improved with PFC
   capacitors.  It requires parallel inductance across the line to
   improve the situation, but this still does not reduce the large
   current flow through the poor ballast cap !

3. A ballast cap in series with the primary, can be "reflected" to
   the secondary side of the xfmr.  It then appears in series
   with the tank capacitor when charging.  This forms a potential
   divider and limits the voltage to which the tank capacitor can
   charge to.  The firing voltage is always less than the supply
   voltage, giving little power throughput.

4. When the spark gap fires,  and the supply is effectively shorted,
   inductive ballasting provides short-term current limiting for the
   HV transformer.  Capacitive ballasting actually results in a big
   current surge through the transformer when the spark gap fires.
   This situation is worse than without any ballasting,  because the
   ballast cap cancels out any little leakage reactance that the
   xfmr has.  Some R or L is always required.

Best reserve PFC caps for what they are designed for:-
Correcting inductively ballasted TC systems !

As a side note,  capacitive ballasting is frequently used in those
little inverter devices for cold cathode backlighting.  For this
application a C is smaller and cheaper than an L !

					Cheers,

					- Richie.