C
Hi, I was hoping for a little advice...
I'm trying to construct an AQL matrix for our QC operators in a flexographic printing outfit. We operate at roughly 3 sigma, which is ~93% yield and fairly standard for our industry.
Our sample size will be 10%, and I need a matrix showing pass/fail levels for different ranges of sample sizes, i.e.
100-500
500-1000
1000-5000
5k-10k, etc.
I'd like to do it in the context of Tim's spreadsheet found in another thread that I can't link to due to forum rules, using an AQL of 7 and UQL of 8. Alpha and beta should be around 5.
My question is, how wide should the sample size ranges be for this to be meaningful? The ISO 2859 standard has huge ranges (35k-150k, 150k-500k) that have the same sample sizes and pass/fail thresholds. It's pretty counter-intuitive to this SSGB that 7 failures in a 150k lot is the same acceptance level as 7 failures in a 500k lot.
I'm feeling very stupid, please help.

I'm trying to construct an AQL matrix for our QC operators in a flexographic printing outfit. We operate at roughly 3 sigma, which is ~93% yield and fairly standard for our industry.
Our sample size will be 10%, and I need a matrix showing pass/fail levels for different ranges of sample sizes, i.e.
100-500
500-1000
1000-5000
5k-10k, etc.
I'd like to do it in the context of Tim's spreadsheet found in another thread that I can't link to due to forum rules, using an AQL of 7 and UQL of 8. Alpha and beta should be around 5.
My question is, how wide should the sample size ranges be for this to be meaningful? The ISO 2859 standard has huge ranges (35k-150k, 150k-500k) that have the same sample sizes and pass/fail thresholds. It's pretty counter-intuitive to this SSGB that 7 failures in a 150k lot is the same acceptance level as 7 failures in a 500k lot.
I'm feeling very stupid, please help.
