kicad-developers team mailing list archive
Mailing list archive
Re: PCBNEW DRC with rotated rectangular pads (partial fix in r1952) [1 Attachment]
Jonas Diemer <diemer@...>
Mon, 7 Sep 2009 08:18:00 +0200
KMail/1.11.4 (Linux/2.6.28-15-generic; KDE/4.2.4; i686; ; )
as I have mentioned last week, DRC Pad2Pad checks are currently incomplete.
Right now, DRC violations between pads that are rotated by an angle not a
multiple of 90° are not detected.
To prevent undetected DRC problems, I think we should be conservative and emit
a DRC error whenever there are rotated pads for which we do not have a DRC
check in place, i.e.
--- pcbnew/drc.cpp (Revision 1954)
+++ pcbnew/drc.cpp (Arbeitskopie)
@@ -1144,6 +1144,7 @@
else // Any other orient
/* TODO : any orient ... */
+ diag = true; // ... until we have fixed this, we assume DRC
What do you think? Should this go into SVN until we have a fix? Who is
currently working on DRC?
Am Freitag, 4. September 2009 17:47:17 schrieb Jonas Diemer:
> Hi everyone,
> find attached a board that generates a false alert in DRC. A "pad near pad"
> error is emitted for every pad of a TQFP-32.
> The problem seems to occur only if the footprint is rotated by 45°+n*90°
> (450, 1350, ...).
> This is due to an incorrect check of orientation in drc.h for the "trivial"
> case that both rectangular pads are oriented by multiple of 90°. The old
> check only tested if both pads were relatively aligned by a multiple of
> 90°, which was also true if both were rotated e.g. by 45°.
> I fixed this in r1952.
> I also noticed that other orientations are NOT CHECKED AT ALL, which
> basically means that DRC errors are not detected for rectangular pads that
> are not aligned to 90°!!! This is kind of bad, although it is marked asa
> Unfortunately, the fix is non-trivial (at least to me)...
> Best regards,
> Yahoo! Groups Links