At 08:21 AM 2/5/2002 +0100, Edi Im Hof wrote:
>Just a quick shot, you said the pads are drawn with an aperture of a 
>quarter of the smallest dimension of the pad. In this case this would be 
>1.25mil. Have you tried to gerber with a 0.01mil resolution? Is such an 
>aperture in the gerber file? If yes, is this aperture actually used?

This is literally on the right track.....

Looking at Mr. Wiseman's files, I discover that there is indeed a zero 
aperture in the RS-274X list. Wondering what the source of this file might 
be, I looked through the PCB file and found two fiducials with arcs with 
zero width, used apparently to keep tracks out. (These arcs should have the 
keepout attribute, but they don't.) It appears that Protel is creating the 
0.0 aperture and then assigning it to the tracks used to draw the rotated pads.

Editing the arc to 1.5 mils draw causes the pads in question to be drawn 
correctly. Deleting the arc also causes the pads to be drawn.

The match width in the aperture setup is set to .00013 mm, which is the 
default. Changing the match width to 0 does not fix the problem, however. 
It appears that there may be a minimum match width, not too surprising. I 
don't have time to check further today, but the match routines could be old 
enough that they allow a 1 mil variation no matter what.

The virtual shorts being used here depend on very precise photoplotting, 
photoplot roundoff can cause difficulties. Normally one creates the pads to 
be integral values in the system being used for plot (that's one problem 
here). Further, rotating the pad causes it to be drawn, introducing other 
possible errors.

Elsewhere I described an alternate method of doing a virtual short, that 
is, a jumper footprint that appears as open to DRC but is actually shorted 
in plot. This alternative involves placing a track, as part of the VS 
footprint, on an otherwise unused mech layer. In the photoplot setups, the 
layer on which one wants the short plotted has a separate photoplot setup 
that merges the appropriate mech layer. This is a tad less simple to set up 
than the virtual short component, but I expect it to be more reliable.

There are several bugs here, interacting. When a zero aperture is assigned 
to draw the pads, they are not drawn, they are missing. The pads themselves 
are not completely drawn even when the aperture is correct, though this may 
be harmless in this case. The zero-aperture arc, however, is drawn. I 
placed a zero-width track and it is also drawn correctly, the problem is 
just in the pad-drawing routine.

Aperture match is not exact even if match is set to 0.

[EMAIL PROTECTED]
Abdulrahman Lomax
Easthampton, Massachusetts USA

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

Reply via email to