--- In
Homebrew_PCBs@yahoogroups.com, "Stefan Trethan"
<stefan_trethan@g...> wrote:
>
> On Thu, 05 Jan 2006 22:18:29 +0100, leon_heller
> <leon.heller@b...> wrote:
>
> >
> >
> > I've uploaded images of the most complex Pulsonix demo before routing
> >
> > and after to Files > LCDpub
> >
> >
> > They are routing1.gif and routing2.gif. Eight layers, 1287 pins.
> >
> >
> > The routing (100%) took 2 minutes 17 secs. I have a got a very fast
> >
> > machine (64-bit twin core Athlon with 1 Gbyte RAM). The router does
> >
> > cost $2,500, though.
> >
> >
> > Leon
>
>
> Well, give me four layers and i route you that mess too ;-)
> I mean that's not really good routing, is it? it's just running traces
> until you hit another one and setting a via.
>
> Also, the parts placement is not good either. It looks like the
parts are
> just put there more random than anything. Some busses seem to run
right
> across the board, much longer than they'd need to.
>
> Now if we ignore these large scale issues (i'm not gonna make 4 layer
> boards with 1287 pins tomorrow), and look at the detail work, it's
still
> crap. i mean, look at the very top very left pad, what is that?
> practically whereever i look i see something i don't like.
>
> This tells me i'm not just too stupid to get my autorouter working
> properly, yours is just as crappy.
>
> Could you try a smaller, single sided board?
I can't see anything wrong with the pad that you mention, perhaps
something was lost when I generated the GIF. I think it's done a
pretty good job. It is about the most sophisticated router available,
at the price. It uses lots of different algorithms to get the job done.
The component placement looks about right to me, The schematic would
be needed to check that.
I don't think you could route that board manually on four layers, and
if you could it would take you a long time.
Autorouters generally don't work at all well on single-sided boards,
they are usually done manually.