Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


SMT line validation

Views: 3797

INGE

#39861

SMT line validation | 22 February, 2006

We have a new SMT line, our customers are asking us the validation of the process. Not only the validation of the resulting assembled board (solder joint...) but also a validation of every single machine in the line (ex. stencil printer...). For example how can I validate a stencil printer? Maybe varying the printing parameters and watching the results (measuring solder paste thickness...)?

Some advice?

Thanks

reply »

samir

#39869

SMT line validation | 22 February, 2006

Is this a FDA type validation?

reply »

INGE

#39872

SMT line validation | 22 February, 2006

Sorry, but FDA is the acronym of what?

Thanks

reply »

#39876

SMT line validation | 22 February, 2006

you could do a capability analysis on the printer.....setup the printer with your print parameters and select a metric (say paste volume deposited) and find out Cp and Cpk values

reply »

Gilligan

#39877

SMT line validation | 22 February, 2006

FDA = Food and Drug Administration. Typically, the FDA requires a validation of every product that will ultimately go to consumers. The validation process is the basic "run it the same way with the same results" over a period of time. Back in my Pharmaceutical days, once a product has been validated for a particular company, the product cannot move to a different location without having to re-validate everything again.

reply »

Dirka Jihabi

#39879

SMT line validation | 22 February, 2006

I worked for a EMS who makes Medical Device Electronics, and yeah, the FDA's methodology for validation is a complete cluster-fudge! They did it on the principle of...does the shop air going into your equipment REALLY 80 PSI...or is there REALLY 220V going into this machine? Useless stuff..That might've worked for manufacturing pharmaceuticals, but not electronics.

This so-called validation method completely ignored actual process capability of the equipment (ie Cpk's for printer, placement accuracy, and oven/profile repeatability)...now THOSE would be useful things to validate in SMT... and of course the resultant defect rate on the product.

reply »


RDR

#39880

SMT line validation | 22 February, 2006

Here are some of the things we had to do when we had to validate process.

Incoming voltage Air pressure squeegee speed placement accuracy oven stability

reply »

#39889

SMT line validation | 22 February, 2006

*Qualify* your equipment by running X boards, Y times, through each piece of equipment under normal operating conditions and quantifying the performance, as in a ppm rating. Then validate your process by defining all of the steps you use to generate the completed products (i.e., Standard Operating Procedures), including how you plan to eliminate any of the defects from getting out into the field.

Your SOP's will include air pressure requirements, operator qualification requirements, step by step operating procedures, parts replenishment, etc., among other things. The deal is, it's up to you to determine what you want to put in there. It's up to them to decide if they're satisfied or not.

I assume you're dealing with a medical device company, so you might want to dig around here:

http://www.fda.gov/cder/guidance/pv.htm

Have fun.

reply »

#39891

SMT line validation | 22 February, 2006

Dirka Jihabi

The FDA doesn't care about the method your company chooses. FDA require in their "Good Manufacturing Practices [GMP]" that your company defines a method and abides by that method. Now, if the fine folk at your company selected a naive method, that's what they chose, not the FDA.

reply »

INGE

#39896

SMT line validation | 23 February, 2006

It isn't an FDA validation. I think Steve has centered the problem. I have already defined all the steps and all the procedures of the process, but now I must run X boards, Y times, through each piece of equipment (as he said before); the question is : How can I quantify the performance of each machine! 1) stencil printer : Run several boards with my process parameters and then measure what? The solder paste thickness (obviously) and then I take an image of the shape of the deposition? Then I must vary the printing parameters in order to estimate the robustness of the process (also worst case) 2)Peak and place : How can I estimate Its performance?

3)Oven : we have a profiler, we can take several profiles and estimate the repeatability

Some advice for these three points?

Thanks

reply »

INGE

#39897

SMT line validation | 23 February, 2006

I forgot: I'm not dealing with a medical device company. But they demand the same one! See you

reply »

#39919

SMT line validation | 23 February, 2006

Defining failure modes will help determine what you need to quantify in all processes. Personally I wouldn't pursue the DoE angle (varying setup parameters to determine robustness).

1)How about % pad coverage, paste thickness, and accuracy? Define what's acceptable (level of performance you need from your printer to achieve soldering that meets your standards), run your samples, verify that all your samples are within your limits of acceptability and you're good to go.

2) Placement accuracy, right parts, right orientation (you'll have to qualify your programming methods somehow as well), etc.

3) Sounds right to me

Is this customer requiring Class 2 or Class 3 standards?

reply »

Dirka

#39920

SMT line validation | 23 February, 2006

Dave F,

I stand corrected. It was my company's naivete (did I spell that right?), and incompetent, inept validation staff.

reply »

INGE

#39979

SMT line validation | 24 February, 2006

If you mean IPC standard classes, they demand class 3.

Bye

reply »

soldering station

Heller Industries Mark 5获得了 2015年中国SMT远见奖 - 回焊炉领域