Following from the EMPF Helpline, prepared by Pat Kane:
If this company is having problems that requires someone to scrub fiducials
with an eraser, then some things could be suggested for them. First, they
should pick a fiducial target that is the easiest for his camera system to
recognize consistantly.  The round .050 solid circle seems to be a very
popular fiducial that most camera systems work well with. This can either
be an exposed HASL circle or one with solder mask over the bare copper.
The mask over copper approach would require that the camera system can
provide varying levels of light to set up the correct recognition
parameter.  The HASL approach has some minor problems with light reflection
that "fool" some camera systems, here again light level changes sometimes
can correct this.  Some camera systems look at the edge
>of the fiducial and ignore the middle, these work well.  I would suggest
>working with the p-p vendor AND the board supplier to find the best
>combination of fiducial finishes and the size most appropriate for his
>machine.

The requestor can contact us directly for further discussion.>
>----------
>>Date: Fri, 24 Jan 1997 21:55:27 +0000
>>To: [log in to unmask]
>>From: Paul Gould <[log in to unmask]>
>>Subject: Re: fiducial recognition problems
>>
>>Hi Ed,
>>
>>Fiducials which are reversed with a laminate centre surrounded by copper
>>or tinned copper seem to be favoured by some but I suppose a lot depends
>>on the type of camera system used. Just a possibly heretical thought; if
>>you are getting consistent pwb's, you could try turning off the fiducial
>>system once you have corrected the programme for the first pwb. This
>>could speed up the process no end.
>>
>>Provided the fabricator is using a good pin registration system for
>>imaging, the main source of variation is in the step and repeat pattern
>>used on the production panel. If each step is identified by a number,
>>the batch can be divided into sets which come from the same step
>>position on the panel. The variation within each set should be much
>>smaller than the variation across all the sets. This should eliminate
>>the need to use fiducial correction for each pwb but only on the first
>>one from each set. Even then the total variation in the pwb's could be
>>much less than you may be getting from the fiducial correction system
>>hence you may be better off without it.
>>
>>I think fiducials are a great idea but we seem to be overdependent on
>>them as a belt and braces or substitute for consistent manufacture and
>>good tooling.
>>
>>Good Luck
>>Paul Gould
>>
>>>A few weeks ago, there was considerable discussion about fiducial
>>>recognition problems on various pieces of equipment and some of the
>>>corrective action that could be taken by the board manufacturer and by
>>>changes in vision system settings.  In my opinion, since we manufacture
>15,
>>>000bds/day, I am not going to have someone sit with an eraser and polish
>>>fiducials.  I need to correct the problem.
>>>
>>>I have seen references to the IPC archive of issues discussed on Technet.
>>>How do I retrieve copies of these discussions so I can find out what was
>>>the final solution?  I would also be interested in a quick re-cap if
>anyone
>>>would care to spend the time
>>>
>>>Thanks
>>>Ed Holton
>>>[log in to unmask]

***ALL NEW EMPF PHONE NUMBERS***

Jack Crawford, HelpLine Coordinator
Electronics Mfg. Productivity Facility
****NEW--317.655.3688
****FAX  317-655-3699
714 N Senate Ave, Suite 100
Indianapolis IN  46202-3112
VISIT OUR HOME PAGE AT: http://www.empf.org
[log in to unmask]


***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to:           *
* [log in to unmask] with <subject: unsubscribe> and no text.        *
***************************************************************************
* If you are having a problem with the IPC TechNet forum please contact   *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask]      *
***************************************************************************