The eye of the storm

[glasseye]

Demo today at 3:50. The code works (finished it last night, and the report), but demos are always tricky. We shall see. My research also progresses at a rapid clip. I need to have a PCB design done in 2 weeks (!). No rest for the weary I suppose...

Comments

[mattyc9.livejournal.com]

What exactly is your code doing?

or

"Please summarize your project in no more than one paragraph"

[glasseye.livejournal.com]

I'm writing an automotive cruise control system for my embedded systems class in C. We're using a small board with an embedded intel 80188ES microprocessor. In this particular lab, we wrote a dynamic round-robin scheduler, drove three LEDs in a simple pattern, used pulse-width modulation to drive a DC motor at varying speeds, wrote status information to the LCD screen (surprisingly difficult since it's a slow 4-bit unit), and measure the period/frequency of an arbitrary square wave input signal. We also used a hardware interrupt driven off of a timer in the CPU for timing.

We had to accomplish all of these things while preserving a scheduler clock cycle of 1 ms. We wrote a simple task to drive an output high when the scheduler was idle. We had about 50% idle time, which was significantly better than several other groups. :)

The webpage is here (http://brl.ee.washington.edu/Education/EE472/), and a pdf of the lab manual (if you're feeling excited) is here (http://brl.ee.washington.edu/Education/EE472/lab/CC_Lab.pdf).

[mbartell.livejournal.com]

ewww cruise is nasty.
I keep thinking about how to do it for real, not always applicable in classes, I know.
are they throwing varied loads (hills and wind gusts) at you too?

I never got far in electronics- more of a hobby. I do remember finding a book one time in the library that had a whole section dedicated to automotive circuits. I copied the solid state sequential turn signals, and had it on my car until the board started to decay, after about 4 years. I need to find the diagram again, so I can do it better.

[glasseye.livejournal.com]

I could've sworn I replied to this comment yesterday.

Anyway yeah, they'll be applying varying loads to the electric motor we're using (this is just a simulation, but from a sytems point of view it acts very similarly to the real deal). The actual cruise control algorithm is pretty straightforward... The real hard part is fitting it all in to this tiny underpowered machine and still have it run with what looks like no latency.

It's quite interesting, though a huge amount of work.