On 11:27 AM 25/05/2001 -0700, Brooks,Bill said:
>I may have had this 'move the polygon fill thing' happen and then hit undo
>to place the polygon back where it was supposed to be... thus causing the
>disconnected image of the fill. I didn't notice it until I went to update
>the fill again later after finishing the changes to the board.. it looked
>like I couldn't change the fill, or delete it. Selecting the entire board
>and moving it all to a new location and allowing the polygons to regen did
>in fact correct the ghosting problem. Its good to have some idea of the
>cause.. I will try to replicate the cause and effect to see how it reacts.
>Thanks for the insight.
>
>Bill Brooks
Bill's basic problem was that the polygon existed on three layers
simultaneously. Yes, that is correct.
Looking at the ASCII it was possible to see that there were tracks on Top,
Mid1 and Bottom all associated with Polygon=1.
Polygon=1 was a small top-layer poly. On the top layer the extent of the
poured tracks was consistent with it's outline. However on the other two
(ghost) layers it was apparently the same size as the full board polygon
that also existed on the top layer. So there were two polygons on the top
layer; Polygon=0 was basically full board and was behaving, Polygon=1 was a
top layer polygon but had associated tracks on three layers. To make
matters worse, the tracks on the two wrong layers covered the whole board
area rather than just the much smaller outline area of the
polygon. Polygon=0 included the area of Polygon=1 - maybe this aspect is a
necessary part of exercising this bug.
I can see how one can get a polygon with tracks on the wrong layer by doing
a Paste|Special and check "paste on current layer" and then say *no* to
repour. This will leave the polygon fill tracks on one layer but the
outline on another. I tried it, it does happen.
Bugs:
1) I fail to see how legitimate user interaction with the database can
cause a polygon to exist on three layers at once.
2) Why/how did the phantom polygon tracks (those on the wrong layers) go
way beyond the extent of the real outline of the polygon. (I could see how
this would occur if the vertices of the poly in question had been edited
way down and then the poly not poured.)
On the PCB in question it looked like the extent of the tracks on the ghost
layers matched that of the good large polygon - was there some buggy
association? Or is this whole thing purely as a result of copying the
polygon onto different layers and editting vertices and not re-pouring? I
think not - I think there has been a corruption somewhere - remember a
single polygon with tracks on three layers.
Finding the small bad polygons outline on the top (correct) layer and
double clicking and re-pouring fixed it. This is why Bill's Select|All and
then move the whole design and then say yes to re-pour (the two) polygons
also removed the ghost tracks.
Personal opinion - I think *not* repouring a polygon when asked should only
be done with great care and knowledge of the likely problems. You should
justify to yourself why you do *not* want to repour rather than the other
way around. I think of repour=yes as the default and no as something to be
used occasionally with care.
I think this issue deserves some comment from Protel CSC as there would
clearly be some bug or sequence of events that can cause these strange
behavior with poly pours.
Bye for now,
Ian
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
*
* To leave this list visit:
* http://www.techservinc.com/protelusers/subscrib.html
* - or email -
* mailto:[EMAIL PROTECTED]?body=leave%20proteledaforum
*
* Contact the list manager:
* mailto:[EMAIL PROTECTED]
*
* Browse or Search previous postings:
* http://www.mail-archive.com/[email protected]
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
