# RE: What size PFC should i use with my 15/30 NST?

```
Terry -

The rearranged equation is still incorrect when used with active volt amps.
The equation is correct only when used with reactive volt amps. To convert
active to reactive amps you need to use complex numbers or trig functions. I
prefer to use trig functions as I show in my post to the Tesla List 7-14-96
"PFC for Neons". Can it be that long ago?

Note that using your "close enough" equation will always give you a leading
power factor which is worst that a lagging power factor. It would be
interesting to see what a model by microSim would show.

To properly meter the TC load you need at least 4 meters, volts, amps,
watts, power factor. The power factor meter is required to tell you if the
load is leading or lagging. A VAR meter would help and save you having to do
the necessary calculations.

Bart I am glad to hear that you are researching the problem. As Terry
pointed out there may be other issues and your work may shed more light on
the subject.

John Couture

------------------------------

-----Original Message-----
From: Tesla list [mailto:tesla-at-pupman-dot-com]
Sent: Friday, September 01, 2000 4:17 PM
To: tesla-at-pupman-dot-com
Subject: RE: What size PFC should i use with my 15/30 NST?

Original poster: Terry Fritz <twftesla-at-uswest-dot-net>

At 09:22 PM 8/31/00 -0700, you wrote:
>
>This equation is incorrect. The Vo and Io are in volt amps. This PFC
>equation is on several web sites and I am amazed it is still being shown.
>You need to find the reactive volt amps (VAR). I showed how to do this in a
>past post that I cannot find now. This is why manufacturers rate their PFC
>caps in VARs.
>
>John Couture
> snip...

Hi John,

This equation is just a rearranged version of a typical VAR
equation.  It also adds line voltage and frequency to make it more
universal.  A simple VAR rating "assumes" voltages and frequencies
which are often different than we would like.  When you short the
output of an NST, The input voltage and current are 90 degrees out
of phase (theoretically).  The reactance Xl = Vi/Ii.

To correct with a PFC cap:

Xl = Xc = 1 / (2 x pi x f x Cpfc) = Vi / Ii

That can be rearranged to:

Cpfc = Ii / (Vi x 2 x pi x f )

However, to make things a bit simpler when using NSTs that are rated by
output voltage and current:

Ii = Vo x Io / Vi

Substituting we get:

Cpfc = Vo x Io / (2 x pi x f x Vi^2)

So it is just a standard equation rearranged specifically for our uses and
using factors with which we are all familiar.  I worked this out one day
and posted it, but apparently it was already well known long before I
reinvented it...

A TC is a pulsed device and such equations are meant for nice steady state
AC waveforms.  The "real" calculations are too nasty for normal humans to
deal with.  However, the above is "close enough" and is simple so that
anyone can use it.  In practice, it tends to give a little lower values than
really needed.  However, having a "little" less PFC capacitance is fine and
probably best.

An alternative is to get ten 20uF PFC caps from DigiKey and add them one by
one to the system until the lowest current is drawn from the AC line.  Or,
do models in MicroSim to find the values.  A true AC power meter and a true
RMS ammeter could also be used to accurately find the power factor, but such
equipment is beyond most people.

Bart Anderson is studying PFC and ballast issues so maybe his work will shed
more light onto all this.  There are probably load situations in the
resonant and Smaller Than Resonant cases where this equation my loose value.
Pig and other external ballast systems may also have unique characteristics.
This is presently being studied...

The equation works fine on my coils and the units in the equation are
balanced.  Is there a better one?

Of course, ideas and comments always welcome...

Cheers,

Terry

```