Choosing a Sampling plan for Start Up Fabless IC Company - Guarantee a 200 DPM limit

G

gbusletta

#1
Hello everyone:
I am new to the world of quality and I am trying to develpo a sampling plan for our ICs.
I have searched this forum and have found quite a bit of usefull information but am still a little confused on choosing a sampling plan.

Here is the situation that I am in:
Our subcontractor does 100% testing of our device.
There is a 200 piece sample (the subcontractor decided on this number) taken from each lot (a lot is typically between 1000-10000 pieces) as a QA sample
If there are any failures in the 200 pieces, the lot is put on hold until we verify the failure(s). If the failure is verified, the defects are replaced and the lot is 100% re-tested.

The questions I have are the following:
1. If I know the sample size is 200, Can I use the binomial distribution, set the acceptance number to 0 (c=0) and search for a LTPD to give me a beta = 10% ( for a 90% confidence interval)? If I do this I get an LTPD = apprx 1.2%.

2. Now if I plot the AOQ, and look for the 1.2% on the x-axis, I get a AOQ of 0.1071% and a AOQL of 0.18%. Can I use either of these numbers to estmate the DPM level of our product? That is if the LTPD is 1.2% and the AOQ is 0.1071%, does this mean that our product is sitting at 1071 DPM? Is it also true that the Maximum DPM level for our product would be 1800 since the AOQL is 0.18%?

3. If all of this is correct, how do I modify the sampling plan in order to guarantee a 200 DPM limit? Do I modify the sample size and the LTPD such that the AOQL is set to 200?

I have read a lot of material on this subject, but it appears a lot of it is repeat, with no real examples to follow.

I also have attached a spreadsheet that I have developed to help me out with this. It is set to the above example. I have set the AQL level to 0 since this appears to be how all of the LTPD sampling plan tables are created.
 

Attachments

Elsmar Forum Sponsor

Tim Folkerts

Trusted Information Resource
#2
g busletta,

A couple of questions for clarification.

Your subcontractor does testing on a sample (200 units)
If those 200 are good, the entire lot is accepted without further testing.
If any defects are identified (and confirmed by you), then you do 100% testing of the entire lot and presumably the entire lot is perfect after that.

Correct so far?

The actual defect rate after testing will depend on the actual defect rate before testing. From your numbers I would say:
  • if you are producing lots at a defect rate of 1.2%, then about 1/10 of the lots will have 1.2% defective and 9/10 will have none defective. The net result is ~0.11% =AOQ, as in your spreadsheet.
  • the worst case would be if you produce lots with 0.5% defective, since that is when you get the maximum AOQ = AOQL = 0.18%.
Thus the best you can guarantee based on just this info would be the AOQL = 0.18% as a long-term average. As you start to see how often you are rejecting lots (and if you count the actual defects in the rejected lots), you should be able to get a better handle on the typical defect rate.

On a similar vein, it looks like you would need a sample of about 1800 to get the AOQL down to 200 PPM.


That said, is there some way to switch from a pass/fail test to a measurement of some continuous variable? Sampling is a good way to check for gross changes in quality, but it is not terribly effective for verifying these small defect rates. For example, rather than simply checking that a voltage is in a particualr range, perhaps you could measure the actual voltage. This will presumably be a more expensive test, but it could probably be done with 30-100 pieces, rather than 1500-2000 to ensure 200 PPM quality.

Tim F
 
D

deniser

#3
Your subcontractor's plan is exactly what we do. I'll explain how we came to that choice and where it will lead you.

Our lot size is variable, 10,000-35,000 pieces. The lot is 100% tested then a QA sample is pulled and retested. We have chosen to use the ANSI Z1.4, normal, sample inspection level II. It refers us to chart M, which refers back to sample size chart L. We've chosen 100 dpm as our goal. To get there, we use 0.065% AQL, single, which results in a sample of 200 pieces, accept on 0, reject on 1. If 90% of our lots pass QA, we can assume from table X-L-1 that the actual outgoing dpm will be .0527% or 527 dpm. If 95% of the lots pass QA the actual dpm will be 256 dpm. If 99% of the lots pass QA, actual outgoing will be 50 dpm.

So long as the lots pass QA at least 95% of the time, your actual defective rate will be at or below 200 dpm. We've seen this in practice many, many times.
 

Tim Folkerts

Trusted Information Resource
#4
Ah!

I think I get it now. The entire lot is "tested" by the subcontractor as basically the final stage of the manufacturing process. Once that is completed, then a sample is drawn to see how well this process of 100% testing actually worked. If the testing process didn't appear to be effective in the sample, then the entire lot is reprocessed through the test equipment.

Out of curiosity, is the sample checked on the same machine that did the initial processing? That could be a problem if the machine itself is at fault. It also seems it would be quite valueable to know how many items were rejected by the initial pass. If this number is unusual, that might suggest checking the original batch AND the test itself.

Have you considered control charting the process? If you check samples during the process, you might catch trouble earlier and not need to reprocess the entire lot through the test stage.

Just a couple more stray thoughts.:)


Tim F
 
D

deniser

#5
You've got it.

We use the same machine for both QA and test. We do a multitude of tester checks before and after both test and QA with a variety of gold standards to make sure the machine is true and correct. We also do real time control charting to ensure it stays good. Monitors are built in all along the process. It's a very impressive system.

Semiconductor test is a slightly different world than most manufacturing test conditions. It is decoupled in both time and space from the rest of manufacturing. You have to know that you're not wasting tester time and the test is giving accurate results.
 
G

gbusletta

#7
Tim:
I answered my own question thanks.
But I have some new ones:
1. Does your comapny do 100% burn in prior to the 100% testing to get remove infant moratlity failures from the population?
2. With regards to rejecting a lot...if a smaple fails from the QA, do you replace the failure and 100% re-test? Do you use the tightened inspection limits etc. if a lot fails?
 

Tim Folkerts

Trusted Information Resource
#8
I don't actually work in a manufacturing environment, so I can't provide answers based on actual experience.

Doing burn in first could be effective. I would suggest a quick calculation of costs. How much does burn in cost and how many defects does it uncover? How much the testing cost and how many defects does it uncover? These numbers ought to give a pretty good idea whether to do burn in first or later.

Tim F
 
G

gbusletta

#9
Hi Tim:
Thanks for all of the info..it has proven to be very useful.
While looking through Z1.4...I am starting to get confused with regards to what they are calling AQL.
All the literature I have seen on AQL define it as a point which, in combination with LTPD (RQL), defines the OC curve.
Now when I go to Z1.4, say for instance to the sampling procedure that we defined in earlier post, AQL = 0.065%, n=200....the charts and graph on Table X-L do not contain the 0.065% in the percent nonconforming. That is if I use the standard of alpha = 5%, the AQL according to what I have read in all other literaure 0.0256% under the header of 0.065% in the standard....do you see my confusion?
I can create the OC curve in excel using the binomial distribution, but I don't understand where the 0.065% comes from...
Can you provide any insight?
 

Tim Folkerts

Trusted Information Resource
#10
I think the problem comes from the definition of AQL and the original construction of the Z1.4 (MIL-STD-105) tables.

Lots with a defect rate equal to the AQL are usually accepted, but the number is not 95%. In fact, "usually" varies depending on the specific sampling plan. It could be up near 99% or it could be down near 90% (i.e. alpha = 0.01 to 0.1), depending on the lot size andthe AQL. The original plans were chosen for convenience, not statisitcal rigor.

If you want statistical rigor, you will need to find (or create) a plan based on specific AQL, LTPD, alpha & beta. If tou stick with Z1.4, you will have to be content with "usually". :(


Tim F
 
Thread starter Similar threads Forum Replies Date
A Getting Stuck Choosing a Sampling Plan Inspection, Prints (Drawings), Testing, Sampling and Related Topics 6
L Choosing the correct Distribution for Acceptance Sampling Inspection, Prints (Drawings), Testing, Sampling and Related Topics 19
S Help needed in choosing the method of calculating the minimum sample size Internal Auditing 12
T Choosing a Notified Body for MDR - SGS/BSI/DNV/DQS-MED Registrars and Notified Bodies 0
E Choosing correct MOPs IEC 60601 - Medical Electrical Equipment Safety Standards Series 2
Q Choosing Nonconformities to Report ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 26
C Quantifying risk in choosing the number of parts, operators and replicates in a GR&R Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 4
Q Old products new class - Dental Devices - Choosing tests EU Medical Device Regulations 2
M Choosing Auditors - ISO 9001 / ISO 27001 (UK) IEC 27001 - Information Security Management Systems (ISMS) 2
T AS9100D - Scope of QMS for New Company - Only Choosing a Function Subset Due to Management AS9100, IAQG, NADCAP and Aerospace related Standards and Requirements 14
E Choosing an ISO 9001 registrar with auditors familiar with our industry Registrars and Notified Bodies 10
M Dilemma about choosing the most applicable clause related to Risk ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 19
J Choosing QMS Software for Aerospace Company Quality Assurance and Compliance Software Tools and Solutions 5
B Choosing not to calibrate (IATF 16949) IATF 16949 - Automotive Quality Systems Standard 6
Q Choosing between ISO 9001 (2015) & TL 9000 certifications ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 4
H On choosing touchscreen displays and ensuring IEC 60601 compliance IEC 60601 - Medical Electrical Equipment Safety Standards Series 6
Q Choosing In-Process Inspection Characteristics Inspection, Prints (Drawings), Testing, Sampling and Related Topics 7
S Choosing a suitable type of Elisa to Test my Sample Misc. Quality Assurance and Business Systems Related Topics 3
S Choosing the correct Elisa Test Food Safety - ISO 22000, HACCP (21 CFR 120) 1
K Choosing a Six Sigma training organization Six Sigma 4
H Choosing between RABQSA Lead Auditor or ASQ CQA Certification Professional Certifications and Degrees 2
S Choosing ISO 9001 Training and if I need the training for work ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 8
S CE Marking choosing between module A, B CE Marking (Conformité Européene) / CB Scheme 2
M Choosing the right Industry Financial Report Manufacturing and Related Processes 1
M Advice for choosing Rechargeable Lithium Batteries to be used in a Medical Device Other Medical Device Related Standards 16
L Choosing a Statistical Test for dissertation results! Statistical Analysis Tools, Techniques and SPC 4
I Choosing a X-Ray Fluorescence Measuring Device for Chrome Coating Inspection, Prints (Drawings), Testing, Sampling and Related Topics 5
J Choosing the Best Standard - ISO 9001 or ISO 13485 ISO 13485:2016 - Medical Device Quality Management Systems 14
B Choosing Inspection Level and AQL AQL - Acceptable Quality Level 1
I Choosing the alpha level in an ANOVA Study Statistical Analysis Tools, Techniques and SPC 3
M Tolerance vs. Study Variation - Choosing the right one Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 5
S Choosing a Higher EER Air Conditioner After Work and Weekend Discussion Topics 7
T Choosing a Mentor Career and Occupation Discussions 3
K Selection Criteria for choosing a Management Representative for QMS ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 13
A Choosing Document Control Software Document Control Systems, Procedures, Forms and Templates 13
S Choosing Parts for Attribute MSA (Measurement System Analysis) in 4th Edition, pg 132 Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 2
C Difference between softwares in choosing correct Taguchi designs Using Minitab Software 16
bio_subbu Choosing a laboratory for biocompatibility test ISO 13485:2016 - Medical Device Quality Management Systems 8
M Choosing a Laser Micrometer - Your recommendations? General Measurement Device and Calibration Topics 8
G Short TQM training - Choosing of most important points Training - Internal, External, Online and Distance Learning 3
B Choosing number of subgroups and subgroup size Statistical Analysis Tools, Techniques and SPC 4
D Choosing a Vision System - Recommendations for a User Friendlier System Quality Tools, Improvement and Analysis 3
D Need Help on choosing dimensional standards / equipment General Measurement Device and Calibration Topics 2
J Choosing Samples for Gage R&R - Randomly picked samples show very little variation Gage R&R (GR&R) and MSA (Measurement Systems Analysis) 13
Tim Folkerts DOE: Choosing a Design - Factors that affect the choice of design The Reading Room 0
B ISO10012:2003 Question - Choosing or assessing the capability of a piece of equipment Other ISO and International Standards and European Regulations 1
gard2372 Choosing between a Small Company or a Large Company Career and Occupation Discussions 17
D Choosing a notified body for a new start up company in the UK - Medical Device area Registrars and Notified Bodies 6
G Choosing Supplier Evaluation Methods - Determining what a Critical Supplier is Supplier Quality Assurance and other Supplier Issues 31
Steve Prevette Choosing Performance Indicators Registered Visitor Articles Archive 0

Similar threads

Top Bottom