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

Re: A belated s.s. realization



Original poster: "K. C. Herrick by way of Terry Fritz <teslalist-at-qwest-dot-net>" <kchdlh-at-juno-dot-com>

Hold on just a minute...

Still having concerns, I made up a drawing comparing my original scheme
with what I've started on as a re-design.  See
http://hot-streamer/temp/prianal.gif.  The drawing shows current flows
for 1/2 cycle of primary excitation.   The circuits do not show the
several primary turns I actually employ but the principle is the same.

Looking at the '"original scheme" circuits, you can see that, during the
transition time when all switches (i.e., transistors) are open (the ckt
to the upper right), inductive-kick current passes through each pair of
low-capacitance coupling capacitors and through fast diodes, clamping the
voltage across the just-turned-off transistors to twice the
electrolytic-capacitor voltage.  At the instant that happens, a resonant
circuit appears consisting of all the presently-conducting inductors (the
primary turns) and all the coupling capacitors.  If the coupling
capacitors have sufficiently low capacitances, the Fr of that circuit at
that time is sufficiently high so that the next zero-crossing of that
current occurs only a modest fraction of time into the next 1/2 cycle of
the transistor excitation (which is at the secondary's Fr).

At that time, the alternate transistors start to conduct (the ones across
the just-finished-conducting diodes) and current starts to flow through
them and the alternate set of inductors (the 2nd set of primary
conductors) in the opposite direction.  The cycle repeats for the
duration of the spark event.

But look at the "N. G." scheme:  Here, it had occurred to me that I might
combine the energy-storage capacitors with the coupling capacitors; make
them one & the same, so to speak.  That's what the drawing shows.  But a
look at the right-hand ckt shows that the Fr during the transition time
is now determined not by small-value coupling capacitors but rather by
large-value electrolytics.  Thus, it will take a very long time, relative
to a secondary-Fr half-cycle, for the "catch"-diodes to quit conducting.
Like, perhaps 1000 times longer.  That is definitely no good: the
alternate transistors will never conduct.

So...does anyone have a comment on this while I go back to the drawing
board?

Ken Herrick

On Thu, 06 Mar 2003 20:09:50 -0700 "Tesla list" <tesla-at-pupman-dot-com>
writes:
 > Original poster: "K. C. Herrick by way of Terry Fritz
 > <teslalist-at-qwest-dot-net>" <kchdlh-at-juno-dot-com>
 >
 > Yeah, I was suffering from a temporary anxiety attack.  As I now
 > realize
 > I'd figured out long ago, my catch-diodes will just do their thing,
 > and
 > when they're finished catching, then the transistors will take
 > over.
 >
 > Ken
 >
 > On Wed, 05 Mar 2003 20:08:05 -0700 "Tesla list" <tesla-at-pupman-dot-com>
 > writes:
 >  > Original poster: "jimmy hynes by way of Terry Fritz
 >  > <teslalist-at-qwest-dot-net>" <chunkyboy86-at-yahoo-dot-com>
 >  >
 >  >
 >  > Hi Ken,
 >  >
 >  > This happens in h-bridge and half bridge circuits too, and is
 > easier
 >  > to
 >  > visualize than in a  "daisy chain". Although it is more evident
 > in
 >  > high
 >  > pulsed power SSTCs, CW SSTCs do this a little bit also. pspice
 >  > really makes
 >  > it easy to understand.
 >  >
 >  > With no load, the current wave form is triangular, and is 90
 > degrees
 >  > out of
 >  > phase with the voltage. During the first few cycles, the current
 > is
 >  > almost
 >  > all positive, so the bottom MOSFET in the half bridge doesnt see
 >  > much
 >  > current. The current starts flowing through the fet near the end
 > of
 >  > the
 >  > cycle.  With no magnetizing current, the current is a sinusoid,
 > and
 >  > is in
 >  > phase with the voltage. At the time of switching, there is no
 >  > current
 >  > flowing, so the diodes dont do anything. In the real case, the
 > catch
 >  > diodes
 >  > carry current for the first part of the cycle, then is carried
 > by
 >  > the
 >  > MOSFETs. The circuit still functions normally; there is still a
 >  > square wave
 >  > voltage on the primary. the increased off center current just
 > heats
 >  > up the
 >  > fets, diodes, and makes the whole thing more confusing. There is
 > no
 >  > reason
 >  > to make any changes if it works :-)
 >  >
 >  > I hope this helps, but it is hard to explain something through
 > only
 >  > words.
 >  > playing around with pspice really helps.
 >  >
 >  >   Tesla list <tesla-at-pupman-dot-com> wrote:
 >  > Original poster: "K. C. Herrick by way of Terry Fritz "
 >  >
 >  > Those of you who have paid attention to my "current-ring" s.s.
 >  > primary
 >  > design may want to read this. As you will recall, I employ two
 > loops
 >  > of
 >  > primary conductor connected together with energy-storage
 >  > capacitors.
 >  > Each loop incorporates multiple switch-transistors and each set
 > of
 >  > transistors acts alternately with the other set to pass current
 > from
 >  > the
 >  > s unidirectionally through first one loop, then the other, thus
 >  > setting up the alternating magnetic field that excites the
 >  > secondary.
 >  >
 >  > Each transistor is protected against reverse--and excess
 >  > forward--voltage
 >  > by the "catch" diode that is connected across its opposite
 >  > transistor, by
 >  > coupling via the associated energy-storage capacitors. But there
 >  > lies
 >  > the rub. It appears that the circuit is a bit like t! he
 > proverbial
 >  > bee:
 >  > unable to fly according to known aerodynamic principles, it
 > flies
 >  > anyway
 >  > since it does not know that.
 >  >
 >  > When one of the sets of transistors is conducting, all is fine:
 >  > current
 >  > flows in a "daisy chain" through the transistors, the primary
 >  > conductors
 >  > and the storage capacitors, setting up that direction of
 >  > primary-current.
 >  > But then, when that set of transistors shuts off at the end of
 > its
 >  > half-cycle and the other set becomes turned on, each transistor
 > of
 >  > the
 >  > other set finds that it is paralleled by a catch-diode that is
 >  > mightily
 >  > conducting in its forward direction, catching the inductive
 >  > overshoot of
 >  > the opposite set of primary conductors. That forward direction
 > of
 >  > diode
 >  > conduction places a low voltage of the wrong polarity across
 > each
 >  > transistor that at that moment is supposed to begin conducting.
 >  > So--those transistors do not begin conducting and the circuit
 >  > doesn't
 >  > "fly".
 >  >
 >  > Except...it does fly; or at least,! is has been doing so. So
 > those
 >  > of you
 >  > who may have been considering use of a similar configuration,
 > take
 >  > heed.
 >  > While I take 5, likely a lot more than 5, trying to figure out
 > a)
 >  > why it
 >  > has worked at all and b) how to change it so that it works
 >  > understandably.
 >  >
 >  > Ken Herrick
 >  >
 >  >
 >  >
 >  > Jimmy
 >  >
 >  >
 >  >
 >  >
 >  >
 >
 >
 >
 >