Of Scopes and CPU's

[Byron G. Jacquot]

>	Since there's all this talk about Scopes and what's proper for
>*analogue* design, what about hybrid designs with CPU's? Of course a
>20-30MHz scope is perfectly acceptable for analogue use... you're not
>going into HF ranges at all, but what about when you're doing digital and
>hybrid designs? How would you pick a good scope for such a thing? Would
>you want to get something that's as fast as the clock frequency of your
>microprocessor? 2x as fast? 

I'm not sure of any hard and fast guidelines here, but I can relate based on
expreience.

For most small microprocessor systems, a 20-30MHz scope will do fairly well.
I've been running various systems in the 10 MHz and down range in the past
couple years, using scopes in the 10-20 MHz range.  While they may not be
the best tools for watching realy high-speed digital stuff, and you can find
yourself on the upper end of the time scale (where the trace can become a
little faint), but it will still work.

I've never found myself thinking "gee...if only this scope went faster."
I've wished for more channels, or probes with finer tips, or for a scope
where I wouldn't leave the input in ground-reference mode, but never
faster...but I might not know what I'm missing.

For low-parts-count microprocessor designs, if you're using sometihng like a
PIC, HC5/11, AVR, lo-speed 8051, etc, etc, you should do fine with a slower
scope.

If you're working with bigger & faster systems (68K, Hitachi SH, etc, etc),
then you'll probably find that a (digital/storage) scope isn't the only tool
you'll need: a memory bus analyzer might also be necessary.

Byron Jacquot