Electron runaway

[jlheidecker]

I have a question about electron runaway, a.k.a. thermionic emission,
a.k.a. electron emission runaway. Does this refer to electron being
emitted by the cathode wires? What makes electrons jump out of the
metal?

Thanks a million Fusioneers,
Jason

[Richard Hester]

The metal is very hot, and the cathode is at a few tens of kilovolts relative to the outer shell. Even if SS or tungsten are not outstanding emitters, they both will emit copiously when white hot and with a few tens of kilovolts bias. The electron emission no doubt causes more ionization in the fill gas, increasing grid temperature via ion bombardment, etc.

[Richard Hull]

Runaway is a vicious and vigorous cycle which folds over rapidly (perhaps exponentially). As the bias increases the gas ionizes a little more vigorously due to field emission. As more and more gas ionizes, the impedance of the device drops, thus allowing heating to begin on the grid which at some point reaches a temperature where it is thermally emissive. This new condition really digs in and the grid temperature rises percipitously. Here is the third real nasty.... When the grid goes white hot, metal atoms are shorn from it and act as a released gas source!! At this point, the pressure inside the system rises rapidly which adds even more heat due to a grossly lowered impedance. I have seen the pressure in the chamber rise from 10 microns to 3 Torr in 2 seconds! This is how much metal vapor can boil off the grid in seconds at the really bad end of a runaway. The cycle will never end until the supply circuit protection or diodes blow or you manually reduce the applied voltage. This is why you must monitor, at minimum, the fusor current and the pressure of the chamber.

Richard Hull

[jlheidecker]

Many thanks to both Richards. So it seems like a bitter-sweet situation...
the thermal emission from the cathode creates more ions for fusion
reactions (a good thing) but can get so out of hand as to destroy itself
(definitely a bad thing). So the big engineering question is how to keep the
thermal emission constant, and at an appropriate level.

[Carl Willis]

Jason,

Ballast or current-limit your power supply, and then don't be too worried about thermionic runaway. In the absence of a current meter, just make sure (if you have a stainless steel cathode anyway) that no parts of the cathode are more than reddish-orange in color. It's easiest to control the system by maintaining constant current at full power, through small adjustments to the vacuum throttling valve. The current responds very promptly; a runaway is immediately alleviated by opening the vacuum valve a bit.

The thermophysical properties of the cathode, its cross-section for ion bombardment, and its work function will determine how much electron current it will provide. The higher the electron current, the more gaseous ionization will occur, hence more ion bombardment to the cathode, hence it gets hotter and emits more electrons, etc. This is the runaway mechanism that must be prevented.

-Carl

[jlheidecker]

> The current responds very promptly; a runaway is immediately
> alleviated by opening the vacuum valve a bit.

This is very good to know. I actually have a conductance control
implemented on my vacuum port which will make it possible to control
flowrate on the fly.

Thanks,
Jason

[ningauble]

What about a tube based constant current source connected to
the ground side of the fusor supply?. You'd need a device
capable of having at least a few KV across it maybe combined
with a limit resistor on the hot side.