Hi Seth,
I have heard this question before and haven't known how
to respond to it. So I have a question for you, if it
could be shown through real data that the board built
from gerber and the board built from ODB are the "same"
(within a certain tolerance) would you be willing to
generate both sets of data and send them out accordingly?
I assume you are comfortable with your system outputting
gerber, it's ODB that has to "prove" it can duplicate the
gerber.
Here's what I have in mind, an experiment. You asked the
question "The problem is how do you insure that both sets
of output data are identical in every way?" Maybe one way
to test that is to test the input and output capabilities
of Genesis/Enterprise.
Let's translate a set of gerber back and forth with ODB a
certain number of times (10?) and see how the result
compares to the "reference data" (first generation gerber).
They would have to pass two tests, 1, a netlist check of
course and 2, a layer by layer feature comparison. The
netlist check is absolute, it has to pass. The features
could not vary more than (1?) percent. The feature
variation test is something we would have to come up with.
Is 1% to small or to large?
Ideally the 10th generation ODB would match the 1st
generation gerber by some agreed upon amount and pass a
netlist compare of course.
I don't know how sound this hare-brained idea is, I thunk
it up driving to work... But I think your question is valid
and has to be answered in some way.
Who would do the testing and presenting of the results?
Would you trust Valor? Frontline? IPC? NEMI? Maybe Valor,
a designer and a fabricator at the very least. Heck, I
volunteer to play if that's what it takes to get more ODB.
Regards,
Mark Steele
CAM Automation Engineer
Toppan Electronics, Inc.
858.695.2222
[log in to unmask]
>-----Original Message-----
>From: Seth Goodman [mailto:[log in to unmask]]
>Sent: Monday, January 28, 2002 6:28 PM
>To: TechNet E-Mail Forum.; Mark Steele
>Subject: RE: [TN] ODB++
>
>
>Hi Mark,
>
>From the design engineer's perspective, supporting two
>different formats
>does create a problem. When you release a design to
>production, it is not
>supposed to be specific to one particular vendor (unless you
>have a captive
>fab shop and NEVER go outside). Unfortunately, not all fab
>shops support
>ODB++. If purchasing would be willing to limit themselves to
>fab shops that
>supported ODB++, that would be fine, but since that limits
>their options it
>is unlikely. So in order to be universal, we would have to
>generate both
>ODB++ and Gerber. This is easy to do but can create subtle
>problems that
>could take lots of time and money to run down. The problem is
>how do you
>insure that both sets of output data are identical in every
>way? While some
>programs allow you to import both data sets and compare, as a practical
>matter, assuring that they are identical is not possible.
>Since there will
>undoubtedly be subtle differences between the two, which one controls?
>Don't we also have to verify and validate both versions by
>building up both
>assemblies? And when problems inevitably arise between boards from two
>different vendors using different data files, who 'ya gonna
>call? Those
>dumb design engineers who created the problem, of course.
***stuff deleted*** (I would like to respond to that in another email.)
>
>Regards,
>
>Seth Goodman
>Goodman Associates, LLC
>tel 608.833.9933
>fax 608.833.9966
>
>
>> -----Original Message-----
>> From: TechNet [mailto:[log in to unmask]]On Behalf Of Mark Steele
>> Sent: Monday, January 28, 2002 4:09 PM
>> To: [log in to unmask]
>> Subject: Re: [TN] ODB++
>>
>>
>> Hi Paul,
>>
>>
>> I work for Toppan Electronics in San Diego, we are a
>> fab house.
>>
>> We strongly prefer ODB data. I tell our Sales people
>> to beg for it!
>>
>> For the past 2 years we have tracked type of data
>> (274D, 274X and ODB) and problems with that data such
>> as bad aperture lists, missing layers, incorrect
>> filename identification, etc, etc, etc.
>>
>> For example, for the last 226 jobs we have input we had
>> 41 issues with gerber, 0 for ODB. That's a lot of phone
>> calls and time spent just getting the data INTO our Genesis
>> system. We allocate 5 min to get the data and import an
>> ODB job, 30-40 for gerber (sometimes it takes hours for
>> gerber, that happens about once a month...)
>>
>> I realize not everyone uses Genesis as their CAM system, but
>> I would wager that a majority of the bigger fab houses do.
>>
>> So, yes, PLEASE generate ODB as a first choice and only send
>> out gerber if asked too...
>>
>> Of course that's just my opinion, I could be wrong...
>>
>>
>>
>> Mark Steele
>> CAM Automation Engineer
>> Toppan Electronics, Inc.
>> 858.695.2222
>> [log in to unmask]
>
---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------
|
