More Reliable Firing at Greater Current

Posted by Rezwan on Apr 06, 2010 at 02:05 PM
Tools: Print | | (4) Comments

At the beginning of month, good shots (those without pre-firing and with pinches, the transfer of energy to the plasmoid and fusion production) were a bit under 50% of the shots we fired.  Since mid-month, we have increased that to 90% good shots.  In addition, we have increased the current FF1 is producing at 24 kV capacitor charge from the 500-600 kA range of February to the 650-850 kA range.  The way we accomplished this was to improve our measurement techniques in accurately adjusting the distance between the electrodes in the spark gap switches and the distance that the trigger electrode (the spark plug) sticks out into the gap.  By equalizing all these distances to within +/- 1%, we have achieved reliable functioning, although all switches still do not fire in complete synchrony yet.

We believe we now know how to fix the remaining switching problems.  First, we found that the use of nickel-alloy automotive park plugs for the trigger electrode in the switches was a really bad idea on the part of the switch supplier.  We had observed that the spark plug tips were melting and re-freezing into rough patterns, which we believe is the cause of pre-firing.  The rough surface provides sharp points for the electric field to concentrate before we can trigger the shot.  To avoid pre-firing, we must now increase the gas pressure in the switches to an extent that some switches do not fire at all.  So, why was the nickel melting?

High-frequency currents, like those in the DPF, are carried only on the surface of a conductor.  It turns out that the depth of this surface layer, the skin depth, is much smaller for ferromagnetic material than for others.  Nickel is such a material.  As a result, the current is too concentrated and heats the surface up to the melting point.

To solve this problem, LPP is producing its own triggering electrodes, made from tungsten-rhenium alloy, a material specifically designed for applications like ours.  We should have these new electrodes produced by May.

In addition, we have developed a highly accurate way of monitoring each switch individually with a fiber-optic detector that converts the light from the switch firing to an electronic signal.  With all of these monitors installed, we will be able to see any late-firing or misfiring switch and be able to adjust it accordingly.  Together, these two steps should enable us to finally achieve full simultaneity of switch firing, which will maximize current output.