Our Micromouse

November 24th, 2005 by sameer | Posted in General, Technical |

While completing my engineering studies, the most gripping and enlightening work I did was the development of an autonomous robot for the Micromouse competition held at Techfest - IIT, Powai.

 

The work spanned almost two years and it helped me see the other side of what was being taught in books.

 How can we be engineers when we have never designed real computer hardware? Ink on paper, the way we learnt to design a 8085, 8086 or 8051 based system, is a pathetic thing to do.

 When I started to design my own 8085 based system, I was horrified to learn how difficult it is. The circuit designed never worked - my best guess being that maybe the speeds of the external  RAM/ROM didn’t match with the speed of the mP and probably should have added some delay elements.

 Watever.

 Finally learnt that in the real world you do not really use a standalone mP and interface it with external RAM, ROM, device driver chips.

You simply use a good microcontroller with onboard RAM and ROM.

Did just that, and it gave life to my first 8051-based system. It worked, and how nicely at that.

Made a test board, with LEDs for each port. Wrote programs, loaded them (for that had to first make an EEPROM programmer) and was delighted to see it working.

 

I wrote a small initialization program which can be called a very scaled down version POST (Power On Self Test) that we see our PCs do when we start them.

Holding a chip in hand, writing a program for it, loading that program into it, and seeing it do what you wanted it to do is comparable to giving birth to a baby. You feel like a creator.

 Designing and testing the peripheral hardware that interfaces with the mC is a joy I can not put in words (and generally can be an extreme hair-tearing frustration as well!)

Interfacing the mC with LPT1 port works fine. I also played around with interfacing it with the PS/2 port. Was delighted to study and understand the AT protocol for keyboard and mouse interfacing.

 Best part was the small venturing into analog circuits, design of infrared sensors.

It’s a great delight to apply theortical knowledge, OPAMPS, infrared LED transmitters and receivers, self designed filter circuits using capacitors, resistances and diodes, and darlington ICs to finally make a sensor that can work at a range of 5 metres reliably. We wanted a shorter range, in the vicinity of 5 cms but had to be extremely reliable. Fine tuning a couple of presets helped us go to a full range of sensitivity levels.

When I look back now at those two years, I miss not being able to do that anymore. Seemed to have moved on to a world of Software without any link to the real world.

Software systems of today alienate people from reality. Someday we will have better systems that will be transparent to us.