Ken,
Im going to say that if anything, the CT is the correct current. I
would personally bypass the 1:1 transformer which is in series with
the CT output for now, just to make sure it is not
interfereing. You are in fact measuring across the 1 ohm, and not
the 1ohm plus the 1:1 Xfmr primary?
The voltage across the shunts, is it a clean sine wave? I could
imagine noise spiking much higher than the real current reading. Or
perhaps the shunts have too much inductance?
Also, keep in mind that with no secondary, the only thing limiting
the primary current in its ring-up is the resistance in the circuit
(which shouldnt be much at all!). So at 40% input voltage, i could
imagine it possible to achieve 1000A pk. How many cycles are you
operating with (i suggest between 5-25 cycles depending on the tank
impedance)? And could you give me your tank C and L values?
Final question, are you using feedback for your drive
signal? Having the gate drive out of tune with the primary LC could
be dangerous.
Well, best of luck in any case. Please get back to me with some
more details as im curious to see what is happening with your circuit.
Steve
On 10/13/05, K. C. Herrick <kchdlh@xxxxxxx> wrote:
Steve (& all others interested)-
An update on my on-going t.c. project (the one with the internal
primary wound in a bucket): I've had to rework, somewhat, my
circuit board, to accomodate TO-220-size transistors for the
coupling transformer drivers and also for the plus-going drivers for
the 4 IGBTs. And, the better to keep the latter off
between sparks, I've replaced the op-isolators that gate them with
darlington 4N30s. Also added some bleeder resistors here & there to
help obviate unpleasant surprises.
After ensuring that the IGBT drive signals are present & properly
phased with no mains-voltage applied, I then briefly applied full
mains voltage with the primary temporarily disconnected, to see if I
have any common-mode IGBT current. I measured none, which implies
that my drivers are working as expected.
So I'm now at the stage of maximum trepidation: starting to pour the
coal to, it so to speak, with the primary connected (but presently
without a secondary). And I find a worry already:
I have 3 means of measuring IGBT current: 1) the 1:100 (cored)
feedback current-transformer in one leg of the primary circuit, its
output connected thru a 1 ohm resistor seriesed with a 1:1
transformer primary whose secondary is shunted by a pair of diodes
back-to-back; 2) a homemade 4 m-ohm resistor in series with the "-"
mains-electrolytics; and 3) a commercial 100 mV/50 A current shunt
in the common bus connecting the "+" and "-" mains-electrolytics.
With mains voltage set by variac at about 40% of max and the circuit
pulsing, I measure, differentially with the scope, approximately a)
10 V peak across that 1 ohm resistor, implying ~1000 A peak in the
primary; b) 25 V pk across the homemade 4 m-ohm resistor, implying
~6250 A peak capacitor current; and c) 25 V peak across the 100
mV/50 A shunt, implying ~12500 A peak capacitor current.
All of the above cannot be true at the same time. Any of the above,
if true, is cause for concern since the mains input is only at ~40%
of max. In an H-bridge, the turned-on IGBTs, the primary circuit
and the "+" and "-" mains-capacitors are all effectively in series,
so I should measure the same current at the 3 places,
more-or-less. But they're significantly, not to say wildly,
different. The IGBTs I am using are Semikron SKM 400GB 123D's.
So...I haven't gotten to pursuing this further yet but it's going to
take some looking-into before I start trying to make sparks. In the
interim, any comments will be appreciated.
Ken Herrick