Yew Jin
7th December 2006, 11:26 PM
We are using DPPM instead of DPMO as our OQA quality indicator. We calculated the DPPM as (total defectives in the sampling size/total sample size) x 1000000. Our lot size is normally fixed at 300.
Are we using the correct method to calculate the DPPM as above?
What is the other quality indicator that we can be used beside the DPPM?:frust:
Miner
7th December 2006, 11:55 PM
If your definition of defective is a complete unit/product, then your calculation of DPPM is correct.
DPMO is typically used when you track defects instead of defectives. Multiple defects are possible per a single unit. This can get difficult unless you set very specific guidelines on how many defect opportunities exist per unit. Many opt for a simple component count of the BOM, saying each component can be good or bad and not trying to define how many different ways it can be bad. DPMO does take part complexity into account, but most customers are just as upset with one bad complex product as they are with one bad simple product.
cuadra
8th December 2006, 12:27 AM
Hi Jin:
I use both (i.e. the different between DPPM and DPMO is just the "O"). Lets look at the definition and it will become clear:
DPMO = 1M * [D / (N*O)]
where:
D= Total number of defects counted in the sample
N= Number of units of product or service
O= Number of opportunities per unit of product or service
1M= One million
Note: Just make the O=1 and your are back to DPPM.
The question is: what do you want to see?
a) The proportion of defective units with respect to your sample
b) The proportion of defects with respect all the opportunities within your sample
In practice: Companies use both for internal analysis but typically disclose only one (DPPM) in order to provide the customer with a number that is easier to understand.
Cuadra
P.S.
DPMO=3.4 corespont to a sigma level of 6, and a yield higher than 99.99%
DPMO=233 -> 5 sigma, yield 99.98%
DMPO=6,210 - 4 sigma, Yield 99.4%
etc
Ernst Kong
8th December 2006, 03:50 AM
The use of Cpk/Cp+Ppk/Pp would likely satisfy your need.
johnnybegood
5th July 2007, 09:09 PM
D= Total number of defects counted in the sample
N= Number of units of product or service
O= Number of opportunities per unit of product or service
1M= One million
How about .... DPPP= D/N
Is this the correct method?