JB wrote:
enough to avoid problems with the transmitter. Multiple series gaps
only
raises the firing voltage, but the breakdown voltage will be set by the
largest gap in the series. A gap at .035 will fire at about 2-10kv
depending on humidity. You only want it wide enough so that it won't
arc
Humidity doesn't change the breakdown voltage very much. In fact,
increasing humidity increases the breakdown voltage. One correction
table for a "rod gap" with 1/2" square electrodes has a 10% correction
going from 15 torr to 30 torr water vapor pressure and a -16%
correction going from 15 to 2.5 torr, which is nowhere near the 5:1
variation cited above. Interestingly, on this basis water vapor is a
better insulator than nitrogen, since density of the humid air is
actually less. Typical breakdown voltage tolerance on a rod gap is +/-
8%.
That was off the top of my head so I stand corrected but there ARE
variations based on other real world stuff like air pressure and dirt on the
insulators. 5-20kv for inside an engine is real world but that depends on
other stuff too, like what kind of spark-plug, compression, resistors,
mixture, timing, but I digress. The point is the gap may have to be
adjusted for best results and might be too wide for protection. I invited
him to read up on the subject because I have other projects I have my head
into

So what are your experiences with spark gap transmission line arresters?
Perhaps you have the precise gap he should be using? My dim recollection is
of a chart for the transmitter power in use.

A good rule of thumb for small gaps with large radius of curvature
electrodes is 30kV/cm (e.g. a 1cm gap between 10cm diameter spheres).

In strongly non uniform field gaps (e.g. needle gaps), the breakdown is
typically 1/3 or less.. that is, 10kV/cm


70kV/inch or 25 kV/inch, respectively.

A sparkplug with a 0.035inch gap would breakdown around 1-2kV in normal
air. In an engine, where the density at firing is probably 5-10 times
higher, the voltage is 10-20kV as you've given.

The usual way to set spark gaps for this kind of thing would be to make
it adjustable, fire up the transmitter at worst case max power and
reflection, and set the gap to be somewhat bigger than causes the gap to
break down. A good worst case would be 2x peak RF voltage (i.e. 2.8
Vrms), since that's what you'd get with a 100% reflection at just the
wrong phase. (unless the source can supply reactive power.. then it's a
lot more complex, because of the potential for resonant rise) You could
use a current limited DC power supply or something like a neon sign
transformer to set the spacing.

But making things practical..

say 2 kW and a 450 ohm transmission line. Erms = sqrt(2000*450) = about
950Vrms.. call it 1350V peak. With sharp ended electrodes (like the
sheet metal V shapes shown in old ARRL handbooks) one might want a
spacing of 0.050 inches?