Subject: | |
From: | |
Date: | Wed, 24 Jan 96 17:19:47 EST |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
The following data is from Circuitest, Nashua, NH:
On 153 boards, they compared the golden board files with files
generated by extracting data ffrom the CAD data base and generating an
IPC-D-356 formatted test file.
Net list errors
No Per cent
Errors 66 43%
Matched 87 57%
Total 153 100%
Basis for errors
No Per cent
Split Plane 90 53%
Revisions 55 33%
CAM 18 10.5%
Signal 6 3.5%
Total 169 100%
I recently took three part numbers and compared the bare board test file
in IPC-D-356 format generated by the fabricator to files for the same revision
boards where the IPC-D-356 test file was generated by an extract routine in
the CAD system. Two matched and one didn't, 66% per cent good and 33% bad test
file data.
I suggest that you extract bare board test files from your CAD system and
supply the test file as part of the data package just like you submit drill
files and Gerber files....IPC-D-350C format contains all three elements...
the drill data, the plot data, and the test net data. We, Digital Equipment
Corporation, supply IPC-D-356 test files with the build package for each and
every build from the first engineering build all the way to production. If the
board fails, you exactly where you stand...fab error, CAM error, film, or
design error. It will pay you back in lost engineering debug and production
testing.
IPC-D-356 test data fromat is also recommended ...it's a bullet proof
format prioven in use. Test service companies such as Circuitest will provide
IPC-D-356 bare board test files from CAD data on a 24 hour turn around.
Harry Parkinson
Digital Equipment Corp
[log in to unmask]
603-884-6760
|
|
|