I take extra care with my prototypes because they not only prove the
concept, but they have to serve as a development tool. In reality,
wire-wrapped circuits are more likely to survive mishandling than a PCB
would be. Whereas you could break a pin or a wire, that's normally quite
apparent to visual inspection. I've got a whole paper-box full of
perfect-looking PCB's which don't work because of some intermittent open or
short due to an overetched or overstressed circuit board.
Where space was a problem, I've been known (rarely) to ship prototypes which
were hand-wired with point-to-point soldered #30 wire. This is much like
wire-wrap, except it require much greater care to cut the wires to just the
right length and strip just the right amount of insulation. It makes for
plenty of wasted wire.
BTW, TechAmerica now sells sheets of 0.100" matrix wire-wrap card with
low-impeance ground/Vcc plane on top/bottom, similar to what I used to build
into my wire-wrap boards. This is really handy, because you can isolate one
section of ground plane from another with a Dremel tool and connect them
with a bypassed ferrite bead or low-value resistor. This makes for
realtively quite regions if you take sufficient pains.
Dick
-----Original Message-----
From: Tony Duell <ard(a)p850ug1.demon.co.uk>
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
Date: Friday, April 02, 1999 11:23 AM
Subject: Re: Kits vs ready-made (was RE: Rebirth of IMSAI)
>
> This strategy is pretty popular with analog work being tried for the
first
> time. I prefer to wire-wrap my digital stuff, and
I've shown that
wire-wrap
Yes, dead bugging takes far too long for most digital stuff. I still
think you need a groundplane for anything with dfast edges (which means
just about anything using modern chips).
> between digital and analog circuitry can be mde to work as well. I
prefer
I've been known to screw a piece of copperclad as a 'dead bug'
groundplane to the edge of a board containing wire-wrapped circuitry. It
might not look all that neat, but it's solid, and it works.
> my prototypes to be really rugged unless there's a specific reason why
one
can't be,
and I'm always afraid to play with those "spiderweb-technology"
circuits, so once built, they never get any improvement from me.
There is rarely any excuse for sloppy construction, even in prototypes.
>
> Now, I've learned a few tricks from those analog weenies who use this
kind
> of stuff all the time. I'll certainly never
forget watching one guy
track
> down a 4MHz clock from a processor getting into
the analog section of his
> circuit by using a divider with points sharp enough to penetrate the
solder
mask and a
spectrum analyzer. He found the section of ground plane which
needed bypass, then fixed it with a couple of capacitors bypassing the
unwanted signal to just the right place. No mean trick if you ask me.
Another trick worth knowing is that sometimes cutting a slot in the
groundplane can isolate a noise source (like that processor) from the
sensitive input circuitry. You want to prevent digital ground currents
from getting anywhere near the analogue side. The impedance of a
groundplane can cause problems.
-tony