New Spark Plugs Pass First Tests

Our new spark plugs, which we hope to be far more rugged than the old, automotive ones, have passed their initial tests.

Production of the spark plugs was slowed during May, partially because of redesigns needed as we went along.  In addition, our first designs, tested in late May, did not work well.  Based on recommendations from our colleagues at Princeton Plasma Physics Laboratory, as well as our own investigation, we had selected Macro, a tough ceramic, as the insulator for the spark plugs.  In addition, we tried to isolate the insulators from the new tungsten pins by leaving a 1/100 inch gap between them.  The tungsten expands slightly, but very rapidly during the current pulse as it is heated by the current.  If the tungsten is in contact with the insulator, the sudden expansion will cause a sharp shock to the insulator.  So the idea was to keep the tungsten and the insulator apart, as well as using an impact-resistant ceramic.

However, this idea did not work.  Even a slight curvature of the tungsten pin caused it to come into contact with the insulator.  In our initial tests, all four insulators broke after two shots.  It seemed possible that more careful alignment might succeed in isolating the pin, but we thought that we needed a much more rugged and reliable solution.  Dr. Subramanian suggested Lexan plastic.  Lexan is extremely hard to break, being used for bullet-proof glass, but melts at low temperatures.  So here we could have a snug fit to the tungsten, since we would not worry about impact, but needed to isolate the insulator for the extremely high temperatures generated at the tip of the tungsten pin by the plasma.  We decide to end the insulator 0.15 inch from the pin tip. In addition, we also would try another plastic, PEEK, which has a much higher melting point, but is somewhat less impact resistant.

After an initial eight shots at 24 kV with three new spark plugs, two using Lexan and one using PEEK, (plus one old spark plug for symmetry) no wear was visible on the insulators, although there was very slight pitting of the tungsten tip.  Since we don’t know how the insulators will react at higher voltages, we have decided to make half the insulators from one plastic and half from the other.  We’re confident this will not affect how fast the switches fire.  We were able to get the three switches we tested to fire with 50 ns of each other, which is acceptable and can be improved with further adjustment of the spark gaps.

We expect to have all 12 spark plugs working by the end of this week.