18 Feb
2010
18 Feb
'10
7:03 p.m.
On Wed, Feb 17, 2010 at 13:49, Tony Duell <ard at p850ug1.demon.co.uk> wrote:
YEs it does!. As I said, it's not the clock rate, it's the switching tiem
and slew rate of the IC. And you can't get 'slow' ICs any more easily.
Even things like GALx cna be touchy. You cna get them to work on a
sodlerless breadboard, but it's not always trivial.
The thing is that experienced electornics hackers know what to look for,
and know what is going to give problems. Beginners don't. They're going
to have problems.
The main problem is that those breadboard are
terrible. It's not the
clock speed that matters, it's the swtiching time of the IC. Most modern
ICs have ouptus that switch so fast that when you combine them with the
stray capacitances on the breadboard and the relatively high impedance
power connections, you get power and ground lines bouncing all over the
place. Without a _good_ 'scope it's impossible to know why your circuit
doesn't work. If you stick to 4000 series CMOS you'll be alright, but
modern 74xxx familes are pushing it. Really pushing it.
But if you're just learning high speed don't matter, even with full
computer prototypes. Incidentally, I just saw this:
http://blog.makezine.com/archive/2010/02/building_a_cpm_68k_computer_from_s…
Crazy, but I'm cheering for him! Retrotastic! :-)
Oh but WHY???
I don't want to discourge people from hacking, and making a 68K computer
(I beleive it's a 68008 processor, given the 48 pin package) is certainly
a fun project. But anyone who is interested enough to want to do that,
and has the abilities to design it, is capable of learning to solder. Period.
My guess is that if you knock the table, at least one connection will
momentarily break, and the thing will crash. Built it on stripboard,
with the same rats nest of wires, and it'll probably work reliably.
-tony