At 05:28 PM 4/2/01 -0700, Kirk Haderlie wrote:
>I placed fills on an internal plane and checked the keepout box.  I assumed
>that the keepout woulds void the copper on the internal plane when gerbers
>were created. It did not.

When you check the keepout box, you are instructing Protel not to plot the 
primitive, but only to use it for the purpose of keeping tracks or other 
copper features out of the area.

However, as Mr. Wilson explained, internal planes are negative layers. To 
remove copper you must plot the primitive. If you uncheck the keepout box, 
you will see that your void is there on the plots. Any non-keepout 
primitive placed on an internal plane will remove copper and is, in that 
sense, already a "keepout."

It would be interesting to experiment and see what the effect of an 
internal plane keep out is on other primitives. If an inner plane "track" 
crosses an internal plane keepout "track," will an error be 
generated?  don't have time to check at the moment.

Protel never dreamed that anyone would put a keepout on an inner plane, 
that is, anyone who knows what an inner plane is.

Inner planes are not DRC'd except to warn where split planes intersect or 
there are other primitives on the plane. This is because checking an inner 
plane is not computationally simple. It can be done -- Wolfgang's tool 
(Router Solutions) can check internal planes -- but it is ... complicated. 
If you are depending on an inner plane for connection, there is presently 
no way -- in Protel -- to verify that a connection has actually been made; 
the connection might be broken if a pad is, for example, surrounded by vias 
close enough together so that the blowouts completely surround the pad with 
no-copper. Or if one has placed free primitives, such as the fill 
mentioned, on the plane layer, that isolate the pad.

This is a major gap in Protel's DRC. Let me hasten to mention that most CAD 
programs likewise do not check plane continuity.

Back in the tape and mylar days, in the last decade, I worked for a large 
company as a design checker; they had a Calma CAD system, in its own temple 
with controlled access, the major part of a million dollars in hardware and 
software *plus* the very substantial overhead. We checked negative planes 
manually; we also drew the layouts manually and they were digitized by the 
priests. I made a suggestion, I don't know if they ever took it: make the 
inner planes by photocombination: dedicate a layer to a tie trace that 
would cross all the pads to be connected (basically an ordinary inner 
layer) and photomerge this with a blown-out negative of the pads. This is 
essentially what I used to do to create a negative plane with tape and 
mylar. Forget cutting little circles in rubylith, which was a bad idea from 
the start (except for creating large blown-out areas on the negative).

I also used this technique to create complex split planes with Tango, which 
did not directly support split planes. It DRCs; one only needs to ensure 
that the tie traces do not cross the split plane boundaries; it's fine if 
they bite into the boundaries, but if they cross into a foreign territory, 
they will cause a short, which would not be detected. This could be 
prevented with Protel by putting keepouts on the routing layer 
corresponding to the negative plane splits.

Protel could greatly improve the efficiency of copper pours and at the same 
time make inner planes DRC-able by implementing a photomerge technique. 
Essentially, all the tools are present; they just need to be linked 
together so that the primitives will be correctly processed. Note that 
RS-274X supports positive/negative merges. I could write more....

Abdulrahman Lomax
P.O. Box 690
El Verano, CA 95433

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To join or leave this list visit:
*                      - or email -
* mailto:[EMAIL PROTECTED]?body=leave%20proteledaforum
* Contact the list manager:
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to