On Sat, 28 Apr 2001 13:02:58 -0700, Brad Velander wrote:

>Terry,
>       from my limited understanding of IPC-D-356 and how it is used at the
>board shop level the issue is not quite as you state below. Using IPC-D-356
>the board shop can check the original netlist (in IPC-D-356 format) against
>the Gerber generation to determine the correctness of the Gerbers as well a
>generating test programs from the netlist.

If that is what they actually do?

The guy who started this thread said 

"I have a request from a board manufacture to generate a Netlist for them
to use for their test lab for the bare boards"

They didn't say if you supply a D-356 netlist will do an additional
verification for you. 

Sounds more like they want to plot the gerbers, make the boards, and test
them against the D-356 netlist you supplied. That saves *them* work, and
puts the onus on you to provide correct gerbers *and* correct D-356 and you
don't get to see any problem till the boards are made and put on test. 

The problem is I don't know if D-356 is adequate for board testing -
marking pads as 'vias' seems optional and is there any facility for net
proximity information to reduce the number of short circuits tests needed.
If the format supports 'intelligent' board testing then an intelligent
netlist is not easy to generate and also depends on the type of board
tester. 

Just to emphasise how important an intelligent test is consider a roving
probe tester doing a PCB with 300 nets. A daisy chain net requires a probe
at each end to verify continuity, some nets have stubs so maybe the average
continuity probes per net is 3 or 4. To verify open circuits between nets
the first net requires 300 probes, the 2nd 299 and so on, an average of 150
probes per net. If you can't cut down the number of open circuit test by
ignoring nets which physically don't come close you can easily increase
test time by a factor of 10. Starting at aroujnd $130k each you really
don't want a board tester wasting 90% of its time (unless the customer is
paying?). 

Does D-356 have proximity information which allows the tester to decide
what it can ignore? Does it have the facility to instruct the tester what
to ignore? Either way generating a good D-356 netlist for test  is not so
trivial. 


Cheers, Terry.


* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
*
* To join or 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]
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to