Re: Triac as gap replacement

["Tesla list" <tesla-at-pupman-dot-com>]

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

Response to Steve and also Rick W:

The turning-off between half-cycles I wouldn't care about since I'd just
want the triacs to stay on during each 120/s damped-wave primary event.
But the di/dt limit seems to be the sticker.  Simulation shows me that,
in my circuit, the di/dt of the first 1/8 cycle (of ~125 KHz) would be
about 2500 A/us.

The extra few volts forward drop of a triac would also be significant
since the rms current would be 20 A or so; an extra volt would mean an
extra 20 W to dissipate.

I do like the notion of triacs (or SCRs) since once they're on, they stay
on.  But I suppose I'll have to give up on them.  I note that the IXYS
VCO132 SCR module, having an Itsm rating of over 3 KA, has a di/dt rating
of only 150 A/us repetitive.  (sigh...)

As to IGBTs...

As I've mentioned previously, I've developed a Marx-like primary circuit.
  Currently (only in simulation), it charges 8 capacitors in parallel from
the mains & discharges them in series, into the primary inductance, via 8
spark gaps.  Because of the way I connect the charging inductors, 160 V
peak mains voltage yields over 600 V capacitor-charge every 1/120 second;
and that yields, from the 8 in series, 4800 V to be applied to the
primary inductance.  That's not too shabby but to realize it I need to
charge the capacitors with the alternating mains-polarity & thus would
need back-to-back transistor circuits to replace each of the 8 spark
gaps.

It seems as if such transistor arrays would end up vastly more expensive
& complicated than spark gaps.  And as to expensive & complicated
transistor arrays..I've been there, I've done that & I am burned out on
it (along with all the late-lamented MOSFETs and IGBTs).

So much, I think, for s.s. gap-replacers.

Ken Herrick

On Thu, 19 Jun 2003 08:12:18 -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>
 >
 > At 20:12 18/06/03 -0600, you wrote:
 > >Original poster: "K. C. Herrick by way of Terry Fritz
 > ><teslalist-at-qwest-dot-net>" <kchdlh-at-juno-dot-com>
 > >
 > >Have any of you looked into triacs having something like these
 > >capabilities?
 >
 > As far as I know, we looked into triacs and SCRs for the OLTC, and
 > rejected
 > them because they take too long to turn on. If you try and turn on a
 > triac
 > with more than a certain di/dt, something horrible happens. For one
 > 25A
 > triac that I looked at, the maximum di/dt was 50A/us.
 >
 > Also, triacs have a pretty high forward voltage drop, at least a
 > couple of
 > volts. This is because they have several P-N junctions in series,
 > none of
 > which can saturate :(
 >
 > The bottom line (I think) is that IGBTs can kick their butts any
 > day. Have
 > a look at a sample triac datasheet to see:
 >
 > http://us.st-dot-com/stonline/books/pdf/docs/7470.pdf
 >
 > Steve C.
 >
 >
 >
 >
 >
 >