What people "should" know

From: Clint Wolff <vaxman_at_qwest.net>
Date: Thu Oct 25 22:14:30 2001

On Thu, 25 Oct 2001, Tony Duell wrote:

> [For the software people here, the best analogy I can think of is the
> GOTO statement (or equivalent). It exists. It's often bad practice to use
> it. But there are times when it's the right thing to use. Just as (IMHO)
> there are times when a monostable is the right thing to use.]

A better analogy for the software people is to assume a variable is
initialized when the program starts. I spent weeks tracking down a
bug which only showed up the second time a program was run. The
debris in memory from the compiler made it happy the first run, but
the leftovers from the first run made the second crash...

One-shots generate a pulse whos timing is based on discrete components,
general a resistor and capacitor, though the resistor may be internal
to the component. With standard components, the resistor is +/- 5% and
the capacitor is +/- 20%. They also change value with age and use,
so one-shots usually are paired with a variable resistor to keep them
tweaked in.

When you are running a board at a few kilohertz, the component
tolerance doesn't hurt you too bad, at a megahertz it's painful,
and at tens of megahertz it just doesn't work. In the good old
days, dedicating 10 flip-flops to generate a pulse was an
expensive answer, now days, (as someone else pointed out)
flip-flops are basically free, but an analog one-shot
(with capacitor) is really huge!

Received on Thu Oct 25 2001 - 22:14:30 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:34:20 BST