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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Process Capability Studies

Verm

#21222

Process Capability Studies | 19 August, 2002

Can anyone lead me to information on capability studies for SMT Production, I know of several points during production to take relevant readings but is there an industry standard to adhere to and if so what?

I have searched the archives to no avial.

reply »

#21233

Process Capability Studies | 19 August, 2002

George: There is no industry standard for the calculation of process capability. Although, you should be aware of IPC-9850, Surface Mount Placement Equipment Characterization

Actually, there is a grand bulk of discussion on process capability in the Archives. Search on �cp*� [without the quote marks]. Two common capability indices are:

Cp. Capability Ratio. Measurement of the width of the distribution of process measurements, compared to a desired point. Cpk. Centered Capability Ratio. Measurement of the mean of process measurements, compared to a desired point.

Capability ratios are a 'free' artifact of using statistical process control. Here�s the process for determining process capability: * Decide on the critical process control parameter; like dwell in wave soldering; though experiments, cause & effect, or FMEA. * Collect data on that process parameter. * Determine the control limits and a center line. * Once control limits are set, inspectors stop the process until out of control limit is addressed and signed of (by operator or tech). * Chart process parameter, selecting the type of chart based on your lot sizing and variable selection. * Determine process capabilities, once your process is under control for any given assembly [and you have progressed to random sampling].

Look to: * My review of "Statistical Process Control for SMT"; William Messina; Datasleuths [Publisher] in the SMTnet Newsletter * Six Sigma for Electronics Design and Manufacturing; Sammy G. Shina; McGraw-Hill; 2002; ISBN 0071395113; 1st edition. I have not seen this book, but Dr. Shina is a professor at University of Massachusetts and a respected presenter of tuitorials at trade shows with alot of industry experience. * Other statistical process control and process capability books referenced in the fine SMTnet Archives.

reply »

Verm

#21237

Process Capability Studies | 20 August, 2002

Cheers Dave.

reply »

EW

#21265

Process Capability Studies | 21 August, 2002

I am running capability evaluations on Dek printers with Cyber and LSM measurements. I spoke with "Phil" last year, but no standard method has been universally adopted. He had some good suggestions. We are currently selecting our own criteria and sticking to it with success. Every defect could arguably be tracked to its original paste volume. We move to placement force later this year. I don't mind sharing or answering questions...this is how I learn more. EWolf@macktech.com

reply »

Bob Willis

#21279

Process Capability Studies | 22 August, 2002

Although the following is not true capability machine assesment I feel that monioring the process at screen print, placement, reflow and wave gives results for both design and process guys to asses thire process. The project has been running since May with updates on the methods used with standards to download at http://www.ppm-monitoring.com

Hope this may be of interest as well

There are no easy methods for companies to compare their manufacturing yields with other similar businesses. This procedure is a necessity to understanding their own world class manufacturing status. A common question asked by many companies is how does my process compare with other companies in terms of yield? The information does not exist or is not easily available to small and medium volume companies.

The lack of process yield data for company comparison is a problem which exists for small, medium and large volume companies and should openly be available to the industry. Many industry and research projects, which receive funding from Europe and UK Government, quote industry issues with little data. Having data on the industry's manufacturing issues will allow future proposals to be targeted correctly.

It is proposed by The SMART Group to create a method by which companies can compare their performance with other companies in similar market places using similar equipment. This would allow each manufacturing process used to build printed board assemblies to be compared. This would initially include screen printing, component placement, reflow and wave soldering and final test.

The most commonly used method of illustrating yield is Part Per Million (PPM) Defective. This provides a measure of the defect level against a process stage or a specific product compared against a known number of opportunities for defects to occur. There are IPC and IEC documents and procedures which do exist to monitor a process, which would be considered during the development phase of this project.

The use of the Internet and e-commerce is an ideal way of gathering and making this information available to the industry in a timely way. Using a combination of website and e-mail would allow access to process information to gauge their own performance. A discussion group or e-mail forum could also be created to discuss issues directly affecting yields and process performance. The same e-mail forum could also be used to circulate the average defect yield each month as well as it being featured on the website. The average PPM levels will be circulated to all other related forums in the industry. Although this project would be co-ordinated by the SMART Group the information would be available to every company in the industry Free.

Although the methods and procedures do exist in a variety of sources, no co-ordinated effort has been made with small and medium volume assembly companies to assist and provide a reference on performance. With support for this project, process data collection can be simplified, a benchmark for each process can be defined and an ongoing reference provided to industry.

The process stages of the project would include:

Organise an initial meeting/workshop of interested parties to define the project and the proposed methods of data collection and gain agreement that this would be a benefit to the industry. This would include industry Groups like the SMART Group, PCIF, ACEM etc. Organisations representing universities and colleges would also be invited, as their expertise on developing and organising data may be useful to the project. Material and equipment suppliers may also be valuable for reference when the basic structure of the project is up and running.

Visit 8-10 companies - these would include high, medium and low volume manufacturers. The aim would be to audit a production line and produce a process performance level for each of the process stages as an initial confirmation of the data gathering procedures. The average yields from the anonymous companies would contribute the first process averages on the main website.

Produce an interim report after 3 months and then develop a website to allow data to be collected and distributed to participating companies. The website would feature the procedures and spreadsheets to allow any company to monitor their own yields and submit the information anonymously to contribute to the average performance figures.

The website would be constructed with a database to allow information to be input by participating companies to calculate the average PPM defect levels for each stage of the process. The average PPM for each process would be available and updated live every month.

When the procedures are defined and the website functioning it would provide a running average and mean of the key stages of the assembly process. Contributing companies to the website database would be able to access data and compare the current process against historic results.

Each process stage would indicate the most common process problems at each stage of manufacture and suggested reasons for the faults. The site would allow registered visitors the ability to download or request documentation and spreadsheets for collecting yield data.

The SMART Group 86 Easton Street High Wycombe Buckinghamshire HP11 1LT England.

Tel:(44) 01494 465217 Fax:(44) 01494 473975 Email: technical@smartgroup.org Web: http://www.ppm-monitoring.com

reply »

Bob Willis

#21280

Process Capability Studies | 22 August, 2002

I would talk to you suppier of placement equipmennt as Assembleon have experience with running odd form on machines as do Panasonic, Siemens, Mimot and Universal.

I have run two lines with machines from your supplier and will be doing a project with some Panasonic equipment next month. If you can put it on your placement machine do it. Even if you machine up carrier plates for your small volume parts.

reply »

dragonslayr

#21290

Process Capability Studies | 22 August, 2002

You wrote: The process stages of the project would include:

Organise an initial meeting/workshop of interested parties to define the project and the proposed methods of data collection and gain agreement that this would be a benefit to the industry. This would include industry Groups like the SMART Group, PCIF, ACEM etc. Organisations representing universities and colleges would also be invited, as their expertise on developing and organising data may be useful to the project. Material and equipment suppliers may also be valuable for reference when the basic structure of the project is up and running.

I would like to suggest getting initial input from the users as well. Industry groups tend to be centric to their area of expertise. Academics have lots of theory, yet can get caught up in their dogma and not consider the practical. (real life)

I have considerable experience, as well as many others, in this subject matter. I do not belong to any group other than "Contract Manufacturers". Yet, in my opinion, this same group would benefit most from the data and experience. Low volume, high mix and the customer wants it yesterday. Not much time to go through the variables one by one and develop the perfect process. Maybe an internet/teleconference with interested end users would be beneficial.

Any comments from the SMTneters at large?

reply »

Jade Series Selective Soldering Machines

Global manufacturing solutions provider