I have spent a fair bit of time trying to `tweak' the rules but seem to reap very 
little improvement. When using 98 we often do several runs using the last run to make 
adjustments to component positions etc until we get a pretty good result that we then 
manually finish off (Have yet to see a Protel finished board that even comes close to 
being `clean' without manual intervention!!!!!).
99 for some reason has consistently taken many hours (and I mean anywhere from 5 to 
20!!!) and not yet given us a single complete route even though 98 does so regularly.

Today, DXP took over 2 hours to route two tracks on a board that has been previously 
routed to about 90% completion - that correlates to about 780 routed tracks out of a 
total of around 840 - and we have given it two empty layers (the pre-routed board is 
4-layers and now we have made it a 6-layer board).

Having said this. Running 99 in my home office (with considerably smaller boards) I 
have no, apparent, problems.

Getting to the stage where we might throw the file at Protel and get them to explain 
the problem. We have just forked out our $9000 for DXP and would expect considerably 
better results than we are seeing at the moment.

Anything on the site that tickles your fancy :-))

Best Regards
Laurie Biddulph
  ----- Original Message ----- 
  From: Bagotronix Tech Support 
  To: Protel EDA Forum 
  Sent: Wednesday, October 29, 2003 7:50 AM
  Subject: Re: [PEDA] Stock Components/Joining nets - now autorouting

  My experience with the Protel autorouter is with Advanced Route 3.1 (Protel
  2.8, 3.X) and the integrated router in 99se.  Don't know about 98 and DXP.

  >From what I have seen, both AR 3.1 and AR 99SE can quickly produce 100%
  results which are acceptable on low to medium-speed digital boards, if the
  routing rules are tweaked just right.  It takes quite a bit of
  experimentation to discover the magic rule set to achieve this.  I have run
  the same board as much as 25 times, tweaking the rules a little each time,
  to get acceptable results.  Fortunately, this is usually feasible since the
  router is so fast.

  All I can think of is that maybe you should tweak the routing rules to
  improve the results.  It's the "cut-and-try" method.

  Laurie, that's an "interesting" website you have there ;-)

  Best regards,
  Ivan Baggett
  Bagotronix Inc.
  website:  www.bagotronix.com

  ----- Original Message -----
  From: "Laurie Biddulph" <[EMAIL PROTECTED]>
  To: "Protel EDA Forum" <[EMAIL PROTECTED]>
  Sent: Tuesday, October 28, 2003 1:45 PM
  Subject: [PEDA] Stock Components/Joining nets - now autorouting

  Thanx everyone for their responses to my queries which were what I expected
  would be the answer.

  On another topic. We have a 4-layer board (nothing unusual mix of analog and
  digital components) which routes on Protel 98 in around 15-30 minutes and
  always achieves 100% routing. If we run the SAME artwork in Protel 99SE it
  takes over 7 hours and only achieves around 90-95% completion. Protel DXP is
  taking around 5 hours and again only 90-95% done. Does anyone have an
  explanation as to why the newer packages are having such a hard time? Are we
  missing some `setup' trick that allows these versions to run as well as
  Protel 98?
  I could understand a longer runtime (say 2-5 hours) if we finished with 100%
  routing and a `good' route (as opposed to one that needs cleaning up like
  Protel 98 usually gives) but so far the results are VERY poor.

  I am running on a Pentium 4 with 256MB RAM and a 1GHz processor so don't see
  that the machine is the problem.

  Best Regards
  Laurie Biddulph

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To leave this list visit:
* http://www.techservinc.com/protelusers/leave.html
* Contact the list manager:
* Forum Guidelines Rules:
* http://www.techservinc.com/protelusers/forumrules.html
* Browse or Search previous postings:
* http://www.mail-archive.com/[EMAIL PROTECTED]
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to