On Thu, 18 May 2006 14:48:43 GMT, John Husvar
wrote:

In article ,
wrote:



A slightly different way of looking at the problem.

The apparent value of a fixed capacitor can be varied by
feeding it through a transformer. For a 1:2 step up transformer
the voltage on the capacitor is doubled so the energy stored in
the capacitor is multiplied by four (CxVsquared).

With a reasonably efficient transformer, measured at the
primary, this appears at the primary as the primary inductance
shunted by a pure capacitor which stores the same amount of
energy. Because this is at half the secondary voltage the
apparent value is four times the value of the secondary
capacitance.

The capacitance value is multiplied by the square of the
turns ratio.

This can be turned into a true variable C by the use of a
Variac to feed the fixed capacitor. The impedance of the
magnetising current taken by the Variac needs to be much higher
than the impedance of the intended capacitor. This is easily
checked by comparing the Variac magnetising current with the
current drawn by the capacitor at a similar voltage. Because
Variacs are usually step down devices the apparent capacitance
value will be reduced by the step down ratio squared.

With a variable capacitor a fixed transformer ratio can be
used but, if 100uF is needed from a 1000pF variable, the values
are pretty extreme - more than 300:1 stepup ratio and, with 200V
in, more than 50Kv across the capacitor!


Jim

I'm stretching long unused knowledge here, but hasn't the OP just
created a big tank circuit to tune the created leg?. The added
inductance of the transformer would tend to even out the current lead of
the caps and balance the created leg's current flow in relation to the
other two.

Or am I totally offbase? (which wouldn't be a first by any means)

This is the reason for specifying that the Variac impedance
be much higher than the capacitor impedance - both of course
measured at supply frequency. So far as the motor is concerned
the the apparent capacitative impedance is the reflected
capacitative impedance paralleled by the inductive impedance of
the variac primary. Since this is a leading power factor paralled
by a lagging power factor this inductance slightly increases the
nett capacitative impedance. This means that the effective
reflected capacitance will be slighly less than the expected
value.

It is of course true that a variac plus capacitor forms a
tuned circuit but, for the conditions specified, this is resonant
at frequency much lower than supply frequency. It is no more and
no less a resonant system than the tuned circuits normally formed
by the motor inductance and the normally added capacitors.

These comments apply to the use of the Variac to trim the
capacitance value - i.e. with the Variac set to the upper half of
its coverage.

If the Variac output is set to a very low value the
impedance of the reflected capacitance may no longer be much
smaller than the Variac primary impedance and this reduces the
value of the apparent reflected capacitance.

The limiting case is when the step down ratio is so large
that the reflected capacitative impedance is equal to the Variac
primary impedance. In this case it is resonant at supply
frequency and the nett reflected capacitance will be zero i.e.
unity power factor.


Jim





















*




--
Posted via NewsDemon.com - Premium Uncensored Newsgroup Service
-------http://www.NewsDemon.com------
Unlimited Access, Anonymous Accounts, Uncensored Broadband Access