TECHNET Archives

May 2014

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Brian Ellis <[log in to unmask]>
Reply To:
TechNet E-Mail Forum <[log in to unmask]>, Brian Ellis <[log in to unmask]>
Date:
Fri, 30 May 2014 09:25:20 +0300
Content-Type:
text/plain
Parts/Attachments:
text/plain (125 lines)
Why? How can you justify such a test?

Two days ago, I wrote in another thread the following which is equally 
relevant to your post:

OK, let me now take it to a more philosophical level. Why do we test? 
Because manufacturing processes are always imperfect. If this were not 
the case, we would not need to test. Testing therefore is really an 
unnecessary, expensive means that prevent us from selling our products 
at a lower price. As, in a real world, testing is necessary, it is our 
duty to minimise testing to a bare minimum, at the lowest cost, that 
ensures the required reliability of the product.

IMHO, it behoves us to question the raison d'ĂȘtre for every test we do 
and, in the relatively rare case where a test is deemed essential, is 
there any way we can reduce its cost impact? Can we rely on statistical 
sampling? If so, as the gaussian bell curve will inevitably show that 
some items in a batch will be out-of-tolerance, can we take the risk? Or 
is gaussian analysis itself acceptable or is the concept faulted?

Now, let's extrapolate this to the subject of this thread by a 
hypothetical example of being an assembly shop receiving a batch of 
1,000 bare boards:

1. Do we need to test them for cleanliness? If no, then don't! If yes, 
then analyse why? Bear in mind that our assembly processes will add 
infinitely more contamination on them than is ever likely to be 
delivered on the bare boards. Will testing ever contribute to the 
reliability of the final product.

2. If testing is deemed necessary, for whatever real or *pretended* 
reason, how do we go about it? We have an infinity of possible tests 
ranging from the lowest-cost PICT (thanks, Graham!), via more costly IC, 
SIR, ECM, electron microscopy, X-ray diffraction etc., etc., to wet 
analyses. Which is the lowest cost method that will give a *reasonable 
insight* into the level, nature and impact of the hypothetical 
contaminants, without going over-the-top? Or would a simple inspection 
suffice? Remember that ANY excessive test is a loss of profits for your 
company and in extreme cases may impact on your own employment if too 
much money is spent on excessive testing, making your product 
non-competitive.

3. Having determined that some form of cleanliness test is required on 
your incoming batch of boards, do you test, say, 1 single sample, 1% 
(i.e. 10), 5% (50) or 100% (1,000)? Are you confident that your sample 
size is truly representative of the batch? Do you say that the whole 
sample of, say, 10 passes, therefore all the remaining 990 will be sure 
to pass? Or do you analyse the 10 results to determine the probability 
of, say, 2 or 3 may fail your pass/fail criterion? If this were the 
case, what would you do about it?

Loads of questions here to answer! IMHO, only when you have answered 
them all, can you make a truly balanced judgement. Consider also that if 
you do two complementary tests, you are not only wasting money but you 
will probably learn nothing ("belt and braces" syndrome!).

I think this is worth more than my traditional 2 eurocents so make it 3 
eurocents (via PayPal) :D

Brian


On 30.05.2014 00:10, Karen Tellefsen wrote:
> We need a 6 hr SIR test for incoming boards.
>
> Karen Tellefsen - Electrical Testing
> Alpha / 109 Corporate Blvd./ S. Plainfield, NJ 07080
> [log in to unmask]
> 908-791-3069
>
>
>
>
> From:   Graham Naisbitt <[log in to unmask]>
> To:     <[log in to unmask]>,
> Date:   05/29/2014 01:34 PM
> Subject:        [TN] PCB Cleanliness testing
> Sent by:        TechNet <[log in to unmask]>
>
>
>
> Hello everyone,
>
> One of you emailed me asking for more information regarding this
> testing?.and I have lost the message - we?ve been having some email issues
>
> here.
>
> Whoever it was, if you might be kind enough to ping me your contact
> details I will be happy to send you additional information
>
> Thanks
>
> Graham Naisbitt
>
> _________________________
> _________________________
> ____________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
> _________________________
> _________________________
> ____________________
>
>
>
>
> ***** Please note that my E-Mail address has changed *****
>
> ____________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please visit http://www.symanteccloud.com
> ____________________________________________________________________
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
> ______________________________________________________________________
>

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________

ATOM RSS1 RSS2