go_away

Author Topic: "real" microcontrollers  (Read 342 times)

0 Members and 1 Guest are viewing this topic.

Offline mstachoTopic starter

  • Supreme Robot
  • *****
  • Posts: 370
  • Helpful? 10
"real" microcontrollers
« on: August 18, 2014, 07:17:00 AM »
This may be a weird question so I figured it should go here:

I've spent most of my time in academia, only really working on industrial projects as a student but never in a full-time industry position.  I'm now working as a lab demonstrator at a university where I'm supporting the mechatronics program, and there is a lot of emphasis on small robotics projects.  Previously, I've used things like the axon and the arduino, as well as the ARM Cortex M-series chips for some special projects.  Currently, we're working with a TI MSP430g2553. 

What strikes me most about the TI chip is its limitations: 16KB flash but only 512B ram.  It's in a 20-pin package so it's hard to find enough pins (or we need to keep switching peripherals on and off to get the labs working).  Oddly, it's a 16-bit uC instead of 8.  It's been weirdly fascinating working with it because it forces me to think REALLY hard about how to use it appropriately.

It got me wondering something: in a project for a company where a uC is necessary, would there ever be a reason to use such a limited chip?  They aren't all that expensive, but neither is, say, a Cortex M0.

For that matter, what does a project using uCs "look like" in an industrial setting?

Just curious,

MIKE
Current project: tactile sensing systems for multifingered robot hands

Offline waltr

  • Supreme Robot
  • *****
  • Posts: 1,944
  • Helpful? 98
Re: "real" microcontrollers
« Reply #1 on: August 18, 2014, 02:03:22 PM »
It varies quite a bit. It could use a PIC12, 8 pin uC or ATiny or it could have a large SOC like a Altera Cyclone with dual ARMs and FPGA.

It all depends on the application and what the engineer chooses.

 


Get Your Ad Here