I’ve made another memristor. This one was fabricated using a method described by Akin (2013), which is a more precise version of the hobbyist method described in this YouTube video. Note: Akin’s method involves heating sulphur powder on a hot plate, which presents a burn or fire risk and is not recommended for amateurs.

I was able to get somewhat stable memristance by making a point contact with a sharpened aluminum leather lacing needle:



This rig was sensitive to vibrations, which is typical of point-contact device rigs. At various contact points and pressures I obtained memristance, open circuits, short circuits, nonlinear resistance, and rectification. The contact would sometimes bounce between these modes in response to light vibration, which was annoying.

I’m experimenting with memristors for two reasons: First, I’m looking for an effective and inexpensive way to provide a laboratory memristor experience for undergraduate students. We’re not there yet, but there do seem to be a few faculty working on this at various universities. Second, I’m following up on my blog post about memristor skepticism. I understand memristance from the circuit perspective; the memristor skeptics tend to be device people (i.e. they focus on fabricating and analyzing the fine-grain physics of individual devices). To the memristor skeptics I would ask: what do you think I made here? If I’m correct in my understanding, I have made a passive device that cannot be reduced to an interconnection of resistors, inductors or capacitors. It is something distinct. If you think memristor theory is a sham, then how should I go about modeling and analyzing this device?

#memristor