[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Best cap size for a sync gap
Original poster: "Malcolm Watts" <m.j.watts-at-massey.ac.nz>
Hi Terry,
On 6 May 2004, at 19:26, Tesla list wrote:
> Original poster: Terry Fritz <teslalist-at-twfpowerelectronics-dot-com>
>
> Hi Malcolm,
>
> Its speculation!! Some NSTs have air gaps and others have the shunts
> crammed in pretty tight metal to metal. Indeed, ones with air gaps
> should not saturate and would theoretically be poor candidates for an
> SLTR coil. But the whole saturating thing is just an idea without
> proof. If one were good, one could find an NST and work out all the
> areas, core types, magnetization curves, ... to see if the idea really
> has a basis in fact. Pretty simple stuff for those that know the art
> of transformer design.
>
> Cheers,
>
> Terry
Well having established that, it might now be possible to make some
real headway into characterizing the transformers in a manner that
could be of some benefit. Laminations that are packed edge - edge are
gapped at least to a minor degree. The way around that in building a
good transformer is to interleave the E's and I's on opposite sides
of the winding for each laminating layer. Any edge-edge gaps that
then exist are rendered a lot less significant since there is now a
significant face - face area between the opposing E's and I's (if I
making this clear ??).
Factors governing NST behaviour might now be arranged in some
kind of order thus:
(1) the fact that the primary and secondary windings are spatially
separated means that a significant amount of leakage inductance will
be present with or without any kind of core.
(2) With no load on the secondary, the shunts will play little if any
part in the scene with the designed-for mains voltage applied (well,
do they? - has anybody calculated or measured whether the x-sectional
area of the core sans shunts is sufficient to keep the magnetizing
current within, say, 1 Amp?).
(3) Again with the rated (or lower) primary voltage applied,
progressive loading on the secondary is going to start forcing some
flux into the shunts, the amount being somewhat proportional to the
reluctance of the magnetic path through the shunts (this includes the
rather large reluctance of any airgap present).
This would be basic behaviour as crudely designed for. Things start
getting complicated once (a) significantly greater than the designed-
for mains voltage is applied to the primary and/or (b), a resonant
circuit is present due to a capacitor hooked across the secondary.
Given the variation between the way different brands of transformers
are constructed, one imagines that modelling starts branching in a
direction dictated by a specific transformer design.
I can easily see where non-linearities would arise without the shunts
and then get a lot more complex with. How closely has any NST been
modelled and quantified with respect to its physical parameters?
Malcolm
>
> At 03:55 PM 5/3/2004, you wrote:
> >Does anyone here have *proof* that the NST shunts are saturating or
> >is this just speculation? The reason I ask is that the shunts
> >actually have an airgap in series with them. I can certainly imagine
> >the main core approaching saturation.
> >
> >Malcolm
> >
> >On 3 May 2004, at 8:20, Tesla list wrote:
> >
> > > Original poster: Kurt Schraner <k.schraner-at-datacomm.ch>
> > >
> > > Hi Gary, Terry, all,
> > >
> > > the subject of best cap size for a sync gap and SLTR is
> > > stimulating, and I like to add my SFr.0.01 to the interesting
> > > thread. Seeking a well sized LTR value for my little coil UBTT
> > > (Uni-Bern-Tesla-Twin) and a reply to Dan McCauley's overvolting
> > > question, I performed 29 SPICE simulations, mentioned in my TCML
> > > posting of last September:
> > >
> > > http://www.pupman-dot-com/listarchives/2003/September/msg00311.html
> > >
> > > The sim's were for just for the _linear case_, where the NST
> > > parameters (measured before) would not dynamically change.
> > > Meanwhile the coil has been realized, and the calculated results,
> > > though in nice agreement with reality otherwise, have taught me a
> > > different lesson: The NST is indeed a more complex animal, than
> > > many of us might have thougt before ;-). I experienced, and fully
> > > agree with those, assigning magnetic shunt/core saturation in
> > > NST's a significant role in TC use.
> > >
> > > The experimental results, compared to the simulation, were as
> > > follows: In simulation a 230V/50Hz line power draw of ~880W /
> > > 1700VA (PF=0.52)at an ~optimal setting of the rotary firing time
> > > of 4.2ms was obtained. The experiment was showing 1300W / 1610VA
> > > (PF=0.807) at a rotary setting of 3.6ms. No humming of the NST was
> > > heard, because of the 1.5m-spark noise. The operation is stable
> > > and reproducible.
> > >
> > > The TC is powered by a 15/60 NST of Italian company FART. The SRSG
> > > motor is fed via a phase-shifter of the John Freau type. The
> > > phase-shifter can also be used to smoothly adjust the power input
> > > to the coil. However I was scared adjusting the phase angle to the
> > > predicted optimal 4.2ms, because of beginning safety-gap firing (2
> > > brass balls of 0.5cm, adjusted ~10mm). A scope-shot of the
> > > charging cycle at lower power may be had at:
> > >
> > > http://home.tiscalinet.ch/m.schraner/UpriC_oszi1304.jpg
> > >
> > > The charging circuit resonance, assuming linear behavior of the
> > > 15/60 NST (L.leak ~795H) and a 33nF primary cap would be about
> > > 31Hz. If saturation brings the leak inductance down to 307H (38%
> > > of linear), the cap would again be resonant! My 1300W operation of
> > > the NST might imply a
> > > leak inductance of about 551H (69% of linear), leading to
> > > 37Hz
> > > resonance - yet away from 50Hz, but the primary LTR cap value only
> > > being ~1.8*Cresonant. Sooo... this alone shows SLTR experimenting
> > > implies some risk, and should be performed cautiously. There
> > > certainly are other effects, interwoven with the magnetically
> > > nonlinear behavior of the NST. More thoughts, or even better:
> > > experimental info, would be very welcomed to be seen appearing on
> > > this great list ;-).
> > >
> > > Data of the little UBTT may be supplied in a later posting, if of
> > > interest. Two spark snapshots are here:
> > >
> > > http://home.tiscalinet.ch/m.schraner/1.06m_1332kl.jpg
> > > http://home.tiscalinet.ch/m.schraner/TeslaAnVorlesung3kl.jpg
> > >
> > > The first pic shows operating the TC in single mode, in my lab;
> > > the second in twin-mode at the Physics Institute of the Bern
> > > University, during a lecture (sparklenghts: 1.06m and 1.3m
> > > respectively).
> > >
> > > Best regards,
> > > Kurt
> > >
> > >
> > >
> > >
> > >
> > >
> > >
>
>
>