Memristor

You read a pie-in-the-sky math paper, and you want a _practical_ answer?

I just brushed up on the memristor in Wikipedia. Take a motor-driven potentiometer of sufficiently low maximum resistance, connect the motor in series with the pot -- voila! A memristor (well, more or less. Or less and less).

My recollection of the memrister was you take a poly resistor and stress it with a voltage. You could change the value of the resistor permanently. The poly resistor was drawn at DR not intended for normal applications.

This was back in the 80's.

Do some more reading..

It does not store energy.. It changes the resistance of the link with applied current on one direction and reverses the value of the R in the other direction of current.

This means if you were to apply AC to it, the value would not change because the alternating currents are shifting back and forth unless for some reason the cycles are interrupted before they hit the base line. This little component could also be used to detect unsymmetrical AC wave forms :)

The idea as far as being used as memory devices would be to do electronic sensing that has very little current in it to detect the value of R and then transfer that as a memory bit 1 or 0. After this read cycle it would then place some current on it to give it a little push back into the direction state it is set for to make sure the reading cycles do not slowly change the value of R over the many reading cycles that can take place.

The device does not require power to maintain the value..

Why don't you google, there is a lot on the subject. Simply put, you're way behind the times.

Jamie

Have you checked out wiki? They explain the details and give several examples.

see Titanium dioxide memristor

Hi,

Last week I meet a german guy from University of Dresden who works with such devices. Chuas paper is not really fresh and the work stikk is in progress.

Marte

Chua's paper was published a long time ago. About 2 years ago, two members of HP Labs created and demonstrated the first semiconductor memristor devices. and published their results -- both the research results as well as a more details-free version (the latter I think in IEEE Spectrum magazine). We at UT Austin created a high-level (SystemC) model of the device, but the results were not al that promising. If you Google for it, you will get the paper references.

What about the Ovonics gadgets? They are just as much memristors as the HP things, which is to say not very much.

HP is hyping things.

John

On a sunny day (Mon, 28 Nov 2011 20:34:43 -0800) it happened John Larkin wrote in :

Those are just CD-RW technology IIRC.

John

On a sunny day (Tue, 29 Nov 2011 03:41:25 -0800) it happened John Larkin wrote in :

OK, I did read it, not so good if soldering erases your non-volatile-memory. That leaves in circuit programming to store the code? Also I do not like the 'heating' idea, as it must be slow to begin with. On top of that 'phase change' may also have a limited number of read/write cycles, as did CD-RW (recordable CDs and DVDs), those also use heating (by a laser, and the change of the properties of the dye in the layer changes the polarisation of the lower power read laser beam.

So, we will see, a thousand new things are invented or discovered every year, very few make it in the market place. I just did read at least 3 useless? inventions today in physics.org. :-)