Date: Wed, 24 Aug 2005 05:33:04 -0500 (CDT)
Subject: had to take board out
X-UID: 82
Content-Type: IMAGE/JPEG; name="img1807.jpg"
Content-Type: IMAGE/JPEG; name="img1808.jpg"

I took the board out. Besides connecting the PC2 and PC3 I/O pins for the 
piezo gyro, all of the unused pins were tied to ground. This is however a 
mistake. All of those wires will have to come off.

Floating CMOS inputs can cause instability. That's what my digital logic 
book says. But the I/O pin on a microcontroller has buffering and pullup 
components connected to it. So the inputs do not float. And if an unused 
pin is tied to ground, then a firmware bug that configures that pin as an 
output and sets it to logical 1 will cause a short.

The other thing I learned is that polarity matters when measuring 
resistances in a circuit. The reason is that there may be a diode 
semiconductor junction between the test points. So when using an ohmmeter, 
you can get different readings when the red (positive) and black (ground) 
test probes switch positions. I had always naively conceptualized 
resistance as a measurement for which polarity does not matter. But this 
is not true. So a general rule when measuring resistance between two 
points is to take two readings for each direction from positive test probe 
to negative test probe.

Two years ago, I set up my machine shop. At that time, every night I did 
some work, I'd learn something new. It was obvious that I didn't know what 
I was doing as I was learning too much. I think that I'm pretty much at 
the same point now with electronics.

There's something about the real world that is too complicated. It's like 
deductive logic fails as a complete model is unavailable. So inductive 
reasoning is required. What I'm talking about is that when working with 
technology, while it is true that theory can correctly indicate the 
correct solution for a problem - the real issue is that you never know 
exactly what problem you are solving.

After work today, I should get the board back in the enclosure so I'll be 
back where I should have been two days ago. And there's no more to wiring 
to do in the electronics box. I think that I'll defer fooling with the 
radios, GPS receivers, IR rangefinders and gyro until later. Most 
important is getting both the computers and microcontroller running.

This is many months off. But before I do anything significant with 
electronics I will purchase an oscilloscope. This is what I'm looking at:

http://www.bitscope.com/product/BS300/

Many months ago I destroyed a servo. I never figured out if it was a 
problem with the servo (bad clock), bad firmware, or some electrical 
problem lengthening the pulse width controlling servo position. For more 
complex mechatronics, a DSO will make it much easier to figure out what is 
going on.