Hi Mark,

We just covered this in my 6 sigma class last week.  You could perform a
binary (Pass/Reject) Inspection Capability Study as part of an Measurement
System Analysis (MSA).

This study requires a minimum of 20 samples with an equal number of
rejectable and acceptable parts.  For each inspector, record the parts
identified correctly, the good parts that were rejected and the bad parts
that were accepted.  This analysis can be performed for each inspector and
the system (all inspectors) as a whole.

Effectiveness (E): inspector or inspection system ability to distinguish
between defective and nondefective parts.

E = Number of Parts Identified Correctly / Total Number of Chances to be
Correct


Probability of False Rejects (FR): likelihood of rejecting a good part

P(FR) = Number of Good Parts Rejected / Total Number of Good Parts Inspected


Probability of False Acceptance (FA): likelihood of accepting a bad part.

P(FA) = Number of Bad Parts Accepted / Total Number of Bad Parts Inspected


Bias (B): inspector or system's tendency to falsely classify a part as good
or bad

B = P(FR) / P(FA)

Bias
B = 1 -> no bias (both errors are equally likely)
B > 1 -> bias toward rejecting good parts
B < 1 -> bias towards accepting bad parts


Inspection Capability Evaluation Guidelines
(These guidelines are weak and should be modified to comply with the
company's stated quality level.  These guidelines state that 90%
effectiveness is acceptable.  Who would admit to knowingly letting 10% of
your inspected product be  incorrectly identified?)

Effectiveness
        Acceptable: ? 0.90
        Marginal: 0.8 - 0.9
        Unacceptable: < 0.8

P(FR)
        Acceptable: ? 0.05
        Marginal: 0.05 - 0.10
        Unacceptable: > 0.10

Probability of False Acceptance P(FA)
        Acceptable: ? 0.02
        Marginal: 0.02 - 0.05
        Unacceptable: > 0.05

Bias
        Acceptable: 0.8 - 1.20
        Marginal: 0.5 - 0.8 or 1.2 - 1.5
        Unacceptable: < 0.5 or > 1.5

Hans

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Hans M. Hinners
Process Engineer
Toppan Electronics, Inc.
770 Miramar Road
San Diego, CA 92126
(858) 695 - 2222 ext. 241
(858) 695 - 6823 fax
[log in to unmask]




-----Original Message-----
From: Mark Orlowski [mailto:[log in to unmask]]
Sent: Monday, June 04, 2001 1:36 PM
To: [log in to unmask]
Subject: Re: [TN] Manual inspection related stuff


Does anyone out there have a lead to any specific research or studies on
the effectiveness of manual inspection of PCBA's?  A crude rule-of-thumb
that we've often used is 80% (80% of actual defects are found by a single
inspection).  I know there's a ton of variables (PCBA complexity, component
technology, magnification, fatigue, etc...) but I would think someone would
have documented a thorough study sometime.  We've got a customer that's
demanding justification for our rule-of-thumb.

Also, can someone recommend a good ergonomics reference regarding manual
inspection?  Thanks in advance.

----------------------------------------------------------------------------
-----
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 the following message: SET
Technet NOMAIL
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
----------------------------------------------------------------------------
-----

---------------------------------------------------------------------------------
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 the following message: SET Technet NOMAIL
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
---------------------------------------------------------------------------------