MEM11A update

Guy Sotomayor ggs at shiresoft.com
Tue Mar 29 22:27:01 CDT 2016


> On Mar 29, 2016, at 5:50 PM, Ethan Dicks <ethan.dicks at gmail.com> wrote:
> 
> On Tue, Mar 29, 2016 at 8:25 PM, Paul Koning <paulkoning at comcast.net> wrote:
>>> On Mar 29, 2016, at 8:18 PM, Guy Sotomayor <ggs at shiresoft.com> wrote:
>>> ...autorouter on Eagle 7.5, so I did this all by “hand”
>>> (at just under 2000 wires it took a while).
>> 
>> I can imagine.  Hand-routing tends to produce much better results;
> 
> Yep.
> 
>> the autorouter in EagleCAD isn't all that good.
> 
> Yep.

I had reasonable luck with it previously but I would route critical signals
by hand and let the autorouter finish everything else.

> 
>> A simple way to find the not yet routed wires is to turn off the metal layers but leave the "ratsnest" layer visible.
> 
> I can suggest triple-checking for unrouted wires.  I have a small PCB
> design I sent out that had *1* unrouted wire segment, between the
> crystal and the pad at the MCU, so short, I couldn't easily spot it
> even after several sessions of looking at it and running ERC.
> Fortunately, there's a handy via _right there_, but each board from
> that run needs a hand-added ECO wire (something I used to do for a
> living 30+ years ago).  v2 is 100% correct!  Lesson learned.
> 

These 3 wires were *really* short.  Nothing I did found them until I came
across a “trick” which is to zoom out so the board is tiny and then with
the route tool active, just “click” on spot on the board.  In short order I
found all 3 remaining wires.  Using that technique, I found the 3 remaining
wires in less than 5 minutes (vs the over an hour looking for them last
night).

Eagle is at least reasonable in that it tells me there are unrouted wires.

Now I have to go through the DRC/ERC and make sure nothing they
complain about is a real problem.

TTFN - Guy



More information about the cctalk mailing list