Corona on the.... GDT??

[nick]

Weird experience I had yesterday. I built a GDT for a full bridge but was testing it on my half bridge. The coil worked as expected (albeit I mostly test at voltages ~45v) but I spent probably 15 minutes trying to figure out where the sound of a second spark breakout was taking place (in addition to the breakout point). I was then surprised to notice that the sound was coming from the GDT (with 2 coils left open) and even more surprised to see a light corona between the wires... it was weird but kind of pretty.

Anyway when I connect it to the full bridge, this effect disappears (unfortunately the coil itself doesn't work but that's another story....).

I'm guessing the disconnected windings are acting like antennas? Interesting that it doesn't seem to negatively affect the performance of the half bridge....

[Weston]

Is your driver circuit grounded? Are you using antenna feedback? The secondary wires are close to the primary windings so the electric field for the two should be ~ the same, so unless things are really wired weird I don't think the unconnected windings are acting like an antenna. If you wanted to eliminate this as a cause you could connect one side of the unconnected GDT winding to a connected winding.

The secondary side of the GDT is referenced to the inverter, which has low impedance paths to defined voltages. Any voltage is likely between the primary and secondary, which means your driver circuit could be floating referenced to the inverter. Are you using a separate power supply for the driver? Is there a common ground or something?