On Wed, 02 May 2001 07:47:42 -0700, Max P. Henzi wrote:

>Hi Max,

>You raise a lot of good questions.  I don't think any of them can be
>answered just "yes" or "no".  My opinion, as the CEO of Lavenir is not so
>relevant. 

Well I considered your opinion might be a little biased but your knowledge
of what goes on in board houses should make it valuable. 

>However, I think it would be worthwhile to give PCB manufacturers
>an opportunity to weigh in on this topic, 

>Terry, perhaps you would like to take the initiative and compose a few
>questions that could be answered in either a "yes" or "no" format or in a
>multiple choice format.  I think there should be no more than five or six
>questions.  Anyone else interested in participating?  Is this worthwhile?

Ouch that sounds like work and I wasn't really that interested, was only 2
minutes ago I discovered P99SE already generates D-356A 'netlists'. That
speaks volumes about current demand for D-356 (and Protel's communication
ability, did no one read what Protel said about D-356 or did they not say
anything?).

I did chat with my board house guy who is a personal friend for a couple of
hours tonight. They are a small shop doing mostly low volumes (10's and
100's not many 1000's) and prototypes. Perhaps they see more different
customers than larger shops. They have a bed of nails tester, a roving
probe tester and Lavinir and other software. 

They have never asked anyone for a D-356 netlist nor so far has anyone
offered them one. 

He saw some benefit in a netlist compare between customer supplied and
gerber derived netlists but thought it would likely throw up far more false
than real errors (Like they add fiducials for the tester and have to assign
a net, or assign nets to unconnected pins for testing, or plating bars on
edge connectors etc). If it cried wolf too much they would not bother
investigating the errors and so no point in comparing in the first place. 

They don't screw up processing at their end often enough for them to
consider it a real problem and when they do it is usually because they get
bad or ambiguous information from their customers. 

If someone could give them a netlist they could load straight into their
tester or make a bed of nails from then how could they complain?, but, he
also thought the chances of that happening were remote. 

He biggest gripe is the quality of gerber and supporting information they
get from some customers. Only around 50% supply enhanced gerber with
embedded apertures which saves them a lot of trouble. They also have
problems Identifying layers and stackups and some packages which require
multiple gerber layers to be merged and they left to guess which should be
negative and how they line up. They are pretty happy to take a job as a
.gck (?) file from GCPreview.

If you want him to fill out a questionnaire it will have to be short and
easy to answer multiple choice questions or it will just go in the bin. 

He (like me) thinks there ought to be a better way but doesn't know what it
is. He also thought they would be working with gerber for a long time to
come because they have so much investment in experience with tools for
gerber. 

I will give him a Protel generated D-356A netlist with my next job, I'm
sure he will be interested, I'm not sure it will buy either of us much. 

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