At 03:23 PM 4/16/2002 -0700, Shuping Lew wrote:
>I tried to load a netlist file to PCB. It has over 1,100 components. I
>receiced a warning of access violation. It says: Access Violation at address
>OF086CC6 module.... Exception Occurred in PCB: Netlist...

First of all, yes, I would strongly suspect a bug, though a damaged 
executable is a possible but unlikely culprit. Mr. Wilson is correct, it 
should not be possible to cause an access violation with bad (or good) 
netlist data.

One factor not stated so far: did the Schematic pass a full ERC?

Many designers, trying ERC, find so many errors and warnings that they do 
not bother to clean it up. After all, the job is due, etc.....

But that quickly-done job is going to be a problem if the PCB contains 
errors that could have been avoided if errors or warnings in the ERC had 
been found and fixed.

Most experienced Protel designers take the trouble to track down and fix 
*every* error and warning. If a warning is verified to come from a 
condition that is intended, a no_ERC directive (Place Directive) is placed 
on the error marker to suppress the report. For example, I'll put such a 
directive on every unconnected pin, because I want to check for unconnected 
pins, for a very high percentage of schematic errors will result in an 
unconnected pin, or a net with no driving pin.

As others have mentioned, a possible cause of the crash is a duplicated 
pin, i.e., a pin which is in more than one net, or possibly a net which is 
found in the list more than once.

The former condition can be created by a bad symbol or by more than one 
part with the same reference designator. The latter will generally be 
detected by ERC. A sign of the former would be that an Update PCB will 
produce macros even after Update has been run twice.

I would not trust that a netlist which crashes Netlist Load will produce 
good results when loaded through the Synchronizer (Update PCB), even if the 
latter does not crash. I would not be content until I had tracked down the 
exact cause of the problem, which can be done by cutting down the net list 
into sections, repeating the process until one has found exactly what 
causes the problem. (Note that a problem might exist because of two 
different nets or node in the list, widely separated in the file, so the 
chunking might need to be done in a more complicated way than merely 
cutting the file in half each time).

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To leave this list visit:
* Contact the list manager:
* Forum Guidelines Rules:
* Browse or Search previous postings:
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to