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

SISG Test Procedure

Original poster: "Mark Dunn" <mdunn@xxxxxxxxxxxx>


I am getting ready to test the SISG concept on a full size coil and just
realized that my test plan is flawed.  I need some help to set up the
test so I don't blow things up out of the box.

The coil is 4.5" OD X 27" lg with a John Freau Toroid resonates at
around 150Khz.  The primary is flat and tapped at about 9 to 10 turns to
resonate similarly with a 36 nF MMC(previously was 38 nF homemade
parallel plate cap).

The important details are that it is uses DC resonant charging with a
bank of (4) MOTS wired parallel(2 anti-parallel) on the primary and the
secondaries in series followed by a full wave rectifier feeding filter
capacitors to reduce ripple to 1000 vdc.  A 31H charging reactor
followed by a blocking diode charges the tank capacitor.

I current limit the MOT power supply with an additional MOT(secondary
removed primary reduced to 60 turns) that is gapped between the I and E
cores.  A .040" gap gives me 16 mH which current limits the power supply
at about 17 amps.

Spark Gap is an RSG running at 460 bps which is about the max allowable
for the power supply to re-charge the cap between bangs.

Here is the problem.  I want to start out with the SISG running at a low
BPS and then bring it up in stages.  This way I can collect data and
analyze as I go.  Specifically, one of the concerns will be IGBT
heating/cooling.  Originally, I thought that I would adjust the current
limiter(by reducing the gap) thus de-rating the power supply and that
this would drop the SISG BPS because the cap would charge slower between

I now think that this is wrong because it is the filter caps that charge
the tank cap thru the charging reactor resonating with the tank cap.
Thus, my plan will cause the process to start with the filter caps fully
charged and the SISG will take-off at 460 BPS.  Since the power supply
will not keep up due to current limiting, the filter caps will gradually
discharge and the BPS will drop until the threshold voltage is not
attainable and the SISG will stop firing.  Nothing like an
unsteady-state test method.

It seems that I need to increase the inductance of the charging
reactor(and leave the current limiter alone).  Problem is a charging
reactor inductance of 100H will only get me down to 150 BPS and where am
I going to get another 70H over my 31H that I have.

Can anyone follow my babbling and suggest an alternative approach that
can get the BPS down for the initial testing without abandoning the DC
resonant charging.


Mark Dunn