> What do you think of my idea of defining board areas where the 
> autorouter can fallback to using alternate (smaller) clearance classes? 



I would say to Alfons that like it.  I could have used it yesterday
and the day before.  But this is with manual routing.  Another way to
solve the problem might be to add another field to a clearance class.
 This field would somehow control which of the two distance values is
used when you are comparing tolerances.  With this field it would
somehow give a higher priority to this class's clearance distance than
to the other object in the test.  So when a track and pin are in the
picture you currently pick the larger of the two clearances.  This
would change that to possibly the smallest.  This solves the track to
pin problem, but it does not solve the track to track problem as you
get nearer to the part.

Dick


>    I appreciate that's work for you, but if you have the time and 
> inclination is it a good solution or is there a better one?   For the 
> moment the best workaround I've come up with is to manually fanout the 
> tracks from the MLF like the legs on a spider.
> 
> Regards,
> 
> Robert.
>


Reply via email to