[sdiy] New FPAA experiments with Scilab/Modelica

[David Riley]

On Nov 28, 2021, at 1:09 PM, Lanterman, Aaron D via Synth-diy <synth-diy at synth-diy.org> wrote:
> 
> Howdy Synth DIYers,
> 
> Lately I’ve been experimenting with a toolset for Field Programmable Analog Arrays based on floating gate technology. My colleague Jennifer Hasler and her team have created a toolset based on Scilab/Xcos in an Ubuntu image you can download and run in VirtualBox. It includes simulation capabilities using Modelica, which is embedded Scilab. I’m documenting some of my progress on YouTube.

Ooh, this is neat.  I spent about 10 years as an FPGA engineer, but I'm always super interested to see how programmable analog has moved along.  It was fairly primitive as of last I checked, but this sounds promising.

> I put together a playlist here:
> 
> https://www.youtube.com/watch?v=4XvQ4LVi0OA&list=PLOunECWxELQTCZEqIbHZpRmIEeQR80o3j
> 
> There’s videos on common source and common drain amplifiers, and voltage dividers that use OTAs instead of resistors (which seems very strange to me, coming from a PCB-level design background, but it turns out that transistors are “cheap” on the chip whereas resistors are expensive in terms of real estate and complicated).
> 
> Jen likes to run her MOSFETs in the subthreshold regime; most books just call this “off,” but it turns out there’s an exponential dependency of current on V_GS that’s reminiscent of BJTs vs. the square law I’m used to with MOSFETs.

I was going to ask if you had any reference for these equations because I'd never been exposed, but I assume this is your colleague?  https://www.youtube.com/watch?v=tNWV4ajz8SI

I'm quite excited to find out more, but if there's any plain reading for me to do, I vastly prefer it to watching videos.  In the absence of that, she's a good lecturer, so the videos are at least quite watchable!


- Dave