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

Re: Rebuilding an OBIT



Original poster: "Jim Lux by way of Terry Fritz <twftesla-at-uswest-dot-net>" <jimlux-at-earthlink-dot-net>

Jochen makes a good observation, and it is indicative of the difference
between an ideal and a real transformer.

In an ideal iron core transformer, all the flux from both windings goes
through the core in the same place, so the flux from the secondary current
DOES counteract the flux from the primary, alleviating the saturation issue.

HOWEVER, in a REAL transformer, the flux linkage isn't perfect (and, in a
NST or OBIT, where the flux leakage is deliberately increased, it's worse),
so there is some component of the primary flux that is NOT cancelled by the
secondary flux, and can lead to saturation.

Also, because of parasitic capacitance in the windings, and their non-zero
resistance, the current phase in primary and secondary is slightly
different, leading to increased NET (i.e. primary-secondary) flux in the core.

But wait, there's more: Flux doesn't travel instantaneously through the
core and there are "iron losses".

The net result is that core cross section IS important, particularly if you
have: non-sinusoidal currents, non resistive power factor, and/or varying
loads.  Hmmmm.. all of these apply to the typical TC supply.

In an NST or OBIT, the core is not only there for "transforming" but also
to serve as the core for a series inductor (the ballast). Not that there is
a physically separate ballast inductor, but in an equivalent circuit sense.

I'd add that this is the primary challenge for designing switching power
supplies: variable load, and it is one reason why the typical PC power
supply needs to have a load on it to work.  These supplies are designed for
absolute minimum cost, which means absolute minimum magnetics, which tends
to limit the current range over which they work (or at least, work
efficiently..)  All of the above factors (leakage L, parasitic R and C,
finite propagation time) are worse at higher frequencies, by the way.  The
electrical design of a switching power supply actually isn't the hard part,
it's the magnetics that is tricky (at least, if you want efficiency, low
cost, or some sense of an optimized design).

Tesla list wrote:
> 
> Original poster: "Jochen Kronjaeger by way of Terry Fritz
<twftesla-at-uswest-dot-net>" <Kronjaeg-at-Stud-Mailer.Uni-Marburg.DE>
> 
> >
> > TRANSFORMERS DO NOT SATURATE DUE TO TOO MUCH CURRENT BEING DRAWN !
> >
> 
> Then how come that you need bigger cores for higer power? Why not just
> thicker wire? Sorry - have to disagree.
> 
> Also, saturation LEADS TO reduced coupling, as part of the magnetic
> flux is no longer confined to the core. This can be tested: the core of
> shorted transformer will attract iron  - you can feel the force if you
> hold a screwdriver close to the core. This is not the case when the
> sec. is open.
> 
> Regards, Jochen
> 
>  ---------------------------------------------------------------------------
>   Jochen Kronjaeger                     Kronjaeg-at-stud-mailer.uni-marburg.de
> 
>   The High Voltage Page  http://www.mathematik.uni-marburg.de/~kronjaeg/hv/
>     info, plans, photos of home-built HV-sources, sparks, HV-experiments
>  ---------------------------------------------------------------------------