Date: Sun, 20 Nov 2005 23:01:38 -0600 (CST)
Subject: steering circuit on motor control board
X-UID: 106
Content-Type: IMAGE/JPEG; name="img2408.jpg"

The steering circuit is on the new motor control board. In the photo, the 
microcontroller is still in the STK500 development board but there are 
three wires running to the sockets where the corresponding microcontroller 
pins would go. It's a relatively fair test this way. Electrically, the 
motor control board doesn't know that the microcontroller isn't on it.

It worked the first time. Testing is done with the small Mabuchi instead 
of the bigger gear motor (from a cordless screwdriver) for the robot. 
There are two reasons.

1. Safety, lower power means lower current which means less chance of
    something catching on fire should there be a mistake.

2. I don't have 4.8 volt power supplies that supply more than 1 amp. Large
    electric motors draw lots of current when first powered.

Layout of components on the punchboard is motivated by minimizing the 
length and crossing of wires. Notice how parts are set back from the edge 
of the board. This makes it easier to handle and run wires around the 
edge. I learned my lesson from the last time.

Two lessons learned recently:

1. Incremental design and testing can not be skipped with hardware. The
    cost of failures is higher than for software. It's not so easy to fix
    bugs. Hardware troubleshooting is very expensive.

2. Power supplies with short circuit detection or a limit on sourcing
    current prevent fires and component destruction. Several times, I've
    become confused or careless and shorted out circuits. Had I been
    testing with batteries or a large industrial power supply, the result
    would have been hardware destruction.