Ferroelectric Memories: The Middle Ground

Our device used a magnet to activate a Hall Magnetic Switch. This magnet was about as strong as your typical Refrigerator Magnet.

I found that when it got close the the MRAM, which was possible in normal operation of the device, the MRAM was destroyed. To be clear it did not just change some bits. It became dead.

I sent these to Everspin and the report I got back was "The base plate was depolarized".

They did update some guidance to avoid magnetic fields after that.

At some point I saw a report that said they were having financial issues (Gee I wonder why, if your devices get erased by refrigerator magnets?). Today no browser will bring up their site:

"You can't visit www.everspin.com right now because its certificate has been revoked."

I've been told that Cyprus Semi, now part of Infineon, devices don't have this issue. I've not yet been willing to try them.

Something else most of this glowing reports, like the parent article, fail to report is that these devices take a lot of current during operation. So if you want to use them in a portable device, like ours, you have to be able to remove their power source and isolate their data pins.

Unless you want to consider an inductor a negative capacitor, we still don't have Negative Capacitors.

That Corp that makes studio grade audio processing ICs does have a "Non-Linear" capacitor that they use in several of their Design Notes:

https://thatcorp.com/design-notes/

Negative Resistance, Capacitance, and Inductance can be synthetized with OpAmp Gyrator circuits.

If you really want to go down into the weeds look up non-optical Parametric Amplifiers.

Here's some of their literature about it: https://www.ti.com/lit/an/slaa526a/slaa526a.pdf

Magnetic-Core Memory : https://en.wikipedia.org/wiki/Magnetic-core_memory