[PEDA] pb short-circuit beetwen pad and internal plane with protel 99se

[Didier Fabbro]

since 5 years this is the third times i have this problem....and always
problem beetwen pad and internal planes
usually i route 4 layers board with power and ground as internal planes
twice i got the 2 internal planes and 1 net connected to a pad !!!
this time it's 1 internal plane and 1 net connected to a pad
the pad in error isn't a special pad....always a pad like others
on the board with same parameters
if i delete this pad and place another pad with same parameters the problem
disappear...!!! the new pad isn't connected to internal plane (as i want)

what i w'll know is why when i have placed this pad on a 24v net, this pad
has been connected at the same time to 3,3v internal plane !!!???

<<You also mentioned that "you can deduct conductor width." I do not
<<understand what this means here. What conductor width?

in relief connection you have the hole and all around like 4 "beans!!"
the thermal fields when the pad is connected to internal plane
conductor width is the copper connection beetwen 2 thermal fields
this is this copper connection which connect the hole to the internal plane


<<Thermal reliefs appear to have conductors, but what is actually
<<plotted is the clearance, not the conducting spokes of the relief.
<<And this pad supposedly is not connected, and therefore shouldn't
<<have any thermal relief, it should have complete clearance.

this pad is supposed not connected ( it's what i want) but it get thermal
relief instead of complete clearance...so it's really connected (it isn't
what i want!!)
if i delete this pad and place another pad with same parameters the problem
disappear...!!! the new pad isn't connected to internal plane (as i want)

<<*Therefore* the problem is either a short *somewhere else* or a
<<photoplot or manufacturing defect. If the manufacturer, for example,
<<used the wrong aperture to photoplot the pads, there you go.

no the problem isn't this!!! the gerber files already are wrong !!!
so the problem came before manufacturing
i have already said that i can see the problem in viewing protel pcb files
this is why i think protel has a bug but i can't get to know what are
the conditions which made this bug !!!

i'm sorry for my english!!!



-----Message d'origine-----
De : PEDA-bounces at techservinc.com
[mailto:PEDA-bounces at techservinc.com]De la part de Abd ul-Rahman Lomax
Envoye : vendredi 14 juillet 2006 21:08
A : Protel EDA Discussion List
Objet : Re: [PEDA] pb short-circuit beetwen pad and internal plane with
protel 99se


At 11:13 AM 7/13/2006, Didier Fabbro wrote:
>for "label attached" i want to say that a net with an assigned name is
>connected to the pad
>yes this is the air gap...i see the same thing with a gerber viewer
>solder mask expansion is different just circular shape
>power plane connect show air gap and you can deduct conductor width
>gerbers aren't differents with gerber viewer like GCpreview

Okay, the gerbers show the pad as being clear from the plane. Just to
be sure, the blowout clearance flash, the circular aperture which
creates the plane clearance, should be larger than the *hole* in
diameter by twice the plane clearance. (which also might be
incorrect, by the way, it might be a special rule for the voltage,
but in the end, if there is clearance on the gerbers, there should be
clearance on the board).

*Therefore* the problem is either a short *somewhere else* or a
photoplot or manufacturing defect. If the manufacturer, for example,
used the wrong aperture to photoplot the pads, there you go.

Now, this raises one possibility. Protel 99SE, as I recall, had a
problem with rotated thermal reliefs. They don't plot properly, they
should never be used. But there should be no thermal relief involved here.

There is something odd about what you wrote. It implies that the
solder plane expansion looks different from the power plane connect A
power plane expansion, to keep the pad from contacting the plane,
will look exactly the same as a solder mask expansion, the only
difference, possibly, would be the clearance. It would seem from the
above that the power plane relief is *not* a circular flash.
(Normally, Protel, as I recall, will plot circular clearances for
non-connected holes, no matter what shape is used for the pad. If one
were, for example, to merge the pads back into the photoplot, on the
theory that having pads on the inner plane layers makes for better
mechanical strength (maybe it does, but it also makes more
opportunity for shorts), this could cause a short. What if the
fabricator did this, and the pad was not round? Or was too large?

Remember, the plane clearance is calculated from the *hole* size, not
from the pad size. Protel assumes, for inner plane generation, that
there is no pad if there is no connection.

You also mentioned that "you can deduct conductor width." I do not
understand what this means here. What conductor width? There has been
no conductor mentioned that is in the power plane, which is
supposedly where the short is taking place. Power planes, including
split planes, are no generated with "conductors" having a width.
Rather, any line primitives placed on that layer are anti-copper, the
"conductor" is what is left after being cut away with whatever is plotted.

Thermal reliefs appear to have conductors, but what is actually
plotted is the clearance, not the conducting spokes of the relief.
And this pad supposedly is not connected, and therefore shouldn't
have any thermal relief, it should have complete clearance.

(But, also, thermal reliefs should never be used to connect what are
essentially vias, not used to solder component pins, to inner planes,
they should always be direct-connect, i.e., nothing is plotted at the
hole location at all, so it connects fully to the plane.)

I know English is difficult for you, but it would be useful if you
write as much as you can, there is no harm if you say more than the
minimum, and it might reveal something about the situation. Explain
what you mean when you write something like "the conductor width."
what conductor, what width, where is is, why is it important, etc.
You know, in trying to describe what you are doing more fully, you
might actually come to realize what the problem is, problem-solving
often works like that.

(You are certainly doing better in English than I would do in French,
or, indeed, in any other language!)



____________________________________________________________
You are subscribed to the PEDA discussion forum

To Post messages:
mailto:PEDA at techservinc.com

Unsubscribe and Other Options:
http://techservinc.com/mailman/listinfo/peda_techservinc.com

Browse or Search Old Archives (2001-2004):
http://www.mail-archive.com/proteledaforum@techservinc.com

Browse or Search Current Archives (2004-Current):
http://www.mail-archive.com/peda@techservinc.com

----------------------------------------------------------------------------
-----------
Orange vous informe que cet  e-mail a ete controle par l'anti-virus mail.
Aucun virus connu a ce jour par nos services n'a ete detecte.