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

Re: Magnetizing current in SSTCs



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

Notwithstanding what I wrote below, I've played around with a basic
simulation of a 1-turn-primary solid-state Tesla coil; a resonant-primary
one.  Its schematic is at http://hot-streamer-dot-com/temp/1-t-pri.pdf.
Might be interesting for anyone with a s.s. driver that will provide just
300 V p-p at Fr and a mere 5 KA of rms current--in bursts, of course,
unless you happen to own a power plant.

One would get the 300 V right off-line via + and - 150 V dc supplies,
then half-bridge transistors to get the drive at Fr.

I found that I could utilize SIMetrix' "ideal transformer" component by
turning it around, i.e., using the primary as the secondary & vice versa.
  I can then fix the (new) secondary's inductance while setting the
"primary:secondary" turns ratio to the reciprocal of the desired one.  In
the simulation case, to 0.002 for the desired 1:500 step-up.  I selected
a coupling coefficient of 0.2 & used a secondary simulation-circuit I
found on the List.

The 7 uF primary-resonator is not too critical: 6-7 makes not much
difference.  C3 (output) current is about 2A.

If I find the motivation to do it and have the simulation components,I'll
try to add to the circuit to make it self-tuning using the secondary's
return-current, & then run a multi-step analysis while varying C3 to find
out how large a change in secondary-circuit Fr it will accomodate.

Comments appreciated...

Ken Herrick

On Tue, 13 May 2003 12:16:00 -0600 "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>
 >
 > Stephen (& all)-
 >
 > It's a good point about the magnetizing current.  To educate myself
 > a bit
 > more, if possible at my age, I simulated a sine wave source, a
 > transformer and a resistive load in SIMetrix.
 >
 > 1.  100 V pk applied at 100 Hz, starting with 0 V at t = 0 and
 > connected
 > via 1 ohm to a "perfect" transformer having 150 mH primary
 > inductance, a
 > 1:1 ratio, the primary: secondary coupling of 1 and a 1 meg load
 > resistor:  1st half-cycle primary current, ~2.0 A pk with the wave
 > becoming symmetrical about 0 A, at ~1.0 A pk, after about 600 ms.
 >
 > 2.  Load resistance reduced to 100 ohms:  1st half-cycle primary
 > current,
 > ~2.5 A pk with the wave becoming symmetrical about 0 A, at ~1.4 A
 > pk,
 > after about 600 ms.  Load current is ~1 A pk.
 >
 > 3.  Same as #1 except coupling = 0.2:  Trivial; essentially the same
 > as
 > #1.
 >
 > 4.  Load resistance reduced to 100 ohms and coupling at 0.2:  1st
 > half-cycle primary current, ~2.0 A pk with the wave becoming
 > symmetrical
 > about 0 A, at ~1.0 A pk, after about 600 ms.  Load current is ~150
 > mA pk.
 >
 > What that tells me is that, a) with a decent transformer, the load
 > current mostly diminishes the magnetizing current; but b) in
 > Tesla-coil
 > work where the coupling is low (like something between #3 and #4),
 > essentially all the magnetizing current is "going to waste" since it
 > is
 > not significantly diminished by the load current...is that right?
 > So
 > those of us who are looking into very low impedance, e.g. 1-turn,
 > primaries are perhaps barking up the wrong tree...right?
 >
 > Further, at 0.2 coupling or so, the reflected secondary impedance is
 > not
 > going to do a whole lot...right?
 >
 > Sadly,
 >
 > Ken Herrick
 >
 > On Mon, 12 May 2003 17:42:10 -0600 "Tesla list" <tesla-at-pupman-dot-com>
 > writes:
 >  > Original poster: "Stephen Conner by way of Terry Fritz
 >  > <teslalist-at-qwest-dot-net>" <steve-at-scopeboy-dot-com>
 >  >
 >  > Here is something for all the SSTCers to think about...
 >  >
 >  > Recently there has been much talk of FBSSTCs,
 > zero-voltage/current
 >  > switching, and so on. Justin & Aron, Jan Wagner and Richie
 > Burnett
 >  > have
 >  > good websites explaining this. What I want to look at is
 > magnetizing
 >  >
 >  > current and how it interacts with these things.
 >  >
 >  > Magnetizing current is the current that would flow in your SSTC
 >  > primary if
 >  > the secondary wasn't there. It is just due to the inductance of
 > the
 >  >
 >  > primary, and so lags the drive voltage by 90 degrees. The fewer
 >  > primary
 >  > turns you use, the bigger the magnetizing current would be. As
 > you
 >  > can
 >  > imagine, the current being out of phase with the voltage messes
 > up
 >  > any
 >  > ZVS/ZCS scheme. The current received wisdom is that this puts a
 >  > lower limit
 >  > on the number of primary turns you can use before your MOSFETs
 >  > cook.
 >  >
 >  > Now, when you add the secondary, the magnetizing inductance is
 > still
 >  > there,
 >  > but the reflected impedance of the secondary appears in parallel
 >  > with it.
 >  > Depending on the drive frequency, this impedance can be
 > inductive,
 >  > resistive, or capacitive. (See http://www.richieburnett.co.uk/
 > for
 >  > nice
 >  > graphs) So here's my point: At a carefully chosen drive
 > frequency
 >  > (it would
 >  > be slightly below the secondary's true resonance) the reflected
 > load
 >  > would
 >  > surely be capacitive and just the right size to cancel out the
 >  > magnetizing
 >  > inductance. Therefore you could use as few primary turns as you
 >  > wanted and
 >  > the current would always be in phase with the voltage.
 >  >
 >  > You can probably make a FBSSTC circuit that runs at this
 > frequency
 >  > automatically. Derive the feedback signal from the primary
 > current
 >  > instead
 >  > of secondary base current or an antenna. This forces the voltage
 > to
 >  > switch
 >  > in phase with the primary current, therefore, the circuit can
 > only
 >  > oscillate at the magic frequency (or in practice probably some
 >  > stupid
 >  > harmonic 8-at-) There is a nice simple half-bridge circuit, used in
 >  > things
 >  > like CFL lamps and electronic halogen transformers, that works
 > like
 >  > this.
 >  >
 >  > Hot or not?
 >  >
 >  > Steve C.
 >  >
 >  >
 >  >
 >  >
 >  >
 >  >
 >  >
 >
 >
 >
 >