Steam-punk lightning arrester for the electric fence charger

This is what the device looked like before I screwed the lid on.

I needed a small project to keep me occupied this afternoon after eldercare. So I whipped up an air-gap surge protector for the electric fence.

Source of data HERE

This data shows the kV required to breakdown an air-gap of various distances of  four different geometries.  The outlier, the blue line, is sphere-to-sphere.  The other three geometries are point-to-sphere, flat end rod-to-sphere and point-to-sphere reverse polarity.

I am pretty sure that my fence charger can withstand 8kV since that is the open circuit output and it would probably be OK with 10kV.  That suggests a 10mm-to-13mm air gap as long as I can avoid having the gap look like sphere-to-sphere.

NOTE: This is only half of the arrestor system.  It also requires a "choke" or inductive load between the fence charger and where the air-gap device attaches to the fence.

Background
Solid metals like iron and copper increase in resistance as they heat up.  The electrical resistance of iron, for instance, increases by about a factor of ten between room temperature and melting.

The resistance of gasses drops to zero for all practical purposes once an arc starts.  Technically, the resistance becomes negative since the voltage between the electrodes DROPS with increasing current due to the increase in free electrons and nuclei.

So an air gap makes a pretty good overload switch.  It has near infinite resistance when it is open and it automatically closes at an appropriate voltage if you have the geometry right.

One advantage of a sphere over a needle or point is that they dissipate heat very well.  Skinny needles don't have much mass and the arc only has one point to leap off of.  Spheres have far more mass and the arc's launch point migrates around the sphere.

This animated gif gives you a sense of how the arc wants to "walk around" if it has a little bit of room.

A compromise is to use a flat ended rod or a cylinder instead of a needle.  The arc nearly always circles around the outermost corner of the cylinder/tube as magnetic forces "push" it sideways, thereby spreading the thermal loading and making this protector more than a one-shot deal.

A picture of my electrodes before trimming.

1/2" rigid copper tubing with 14 gauge copper wire soldered in for leads.

This is what the back end looks like.

The whiskers on the business end were trimmed short, about 1.5mm by my calipers.  I want to give the field a very clear "hint" about where they should initiate the arc.  To wit, I want to avoid the perfect field effect that forces the breakdown voltage to 22 kV.

Dry assembled.  The white is some PVC pipe I needed to shim the copper pipe to fit the holes in the conduit body.

Dry assembled.  I did not want the two initiators 180 degrees apart but I did not want the to line up either, so they are about 90 degrees apart.

Why dry assemble?  To get the gap close before attaching the conduit box to a mounting board.

Screwed the 1/2" conduit body to the mounting board and "glued" the other parts in place with silicon adhesive.  9.5mm from the tip of the initiator to the edge of the tubing opposite it.

A close-up showing the silicon squeeze-out.

One nice thing about using "electrical" boxes is that they will help contain the arc and minimize the risk of starting a fire.  Another advantage is that copper parts in an enclosed electrical box stay clean!

Added later:  I know that some of my readers have more electrical design expertise in their left pinkie toe than I have in my whole brain.  Let me know if I am missing anything important or if I am way overthinking the issue.