Jim, where I work we have done some SPC work.  Some has been for training 
and demonstration, with the hope that  benefits would come from it.

Here are some places we are using SPC.

Electroless copper deposition (has been forcing engineer to look at data 
operators have been collecting)

Hot Air Level post-clean (has allowed reduction in frequency of changes 
of cleaning bath)

HAL solder contamination (copper, has put some discipline into the 
contamination testing)

Etch before electroplating (has forced engineer to look at the etch bath 
and adjust it to a better operating point, saving energy and materials)

Clean room particle counts (detected cases where someone had shut off the 
air filtering mechanism)

Daily adjustment of photoplotter (not sure how much good this can do, may 
detect degradation of mechanicals)

Wastewater treatment

Developer pH (can detect cases where developer has been tinkered with)

There are a lot more things we hope to work on.  One thing I have 
observed, SPC may be helping to keep people from responding to false 
alarms.  

I would appreciate comments from any and all on the subject of SPC.  

The section in PC-90 on measurement quality, repeatability and 
reproducibility, is very interesting.  We have been looking at some of 
our measurement methods and thinking of ways to improve them.

Lou Hart [log in to unmask]


.

On Wed, 6 Mar 1996, Jim Douglas wrote:

> this message is aimed at board shops!
>   which departments/processes do you carry out spc?
>   we are looking for some more really "meaty" areas to carry out spc
>   ie. areas where we can get some real benefit and not just put up charts for the sake of it,or to confuse an auditor.
>   note we already have a copy of  IPC-PC-90
> 
>                                           JIM DOUGLAS
>                                                KAM CIRCUITS
>                                     phone   00441249815262
>                                      fax        00441249814726
>                                      email      [log in to unmask]
>