How does one accomplish this conversion?

Formual?

JimDad

Motorola has a book with tables.

sigmas DPMO(npmo) non-conformities or ppm
1 317,400
2 45,400
3 2,700
4 63
5 0.57
6 0.002
7 0.000003

This data above doesn't include a 1.5 sigma shift on either side of the median.

The book title is The Nature of 6 Sigma Quality from the Motorola University Press.

Try the following site for a simple converter: http://www.pyzdek.com/Sigma-PPM-Percent%20Converter.htm

Thanks for the feedback folks, I really need to find the formula. Unfortunately, the table does not produce a sigma for a specific DPMO and I do not have the entire 2000 office suite to utilize Pyzdek's converte page.

Still seeking directions,,
Jim

Folks, I am deserately seeking a formula for converting DMPO to sigma...

Anyone??

Thank you.

Here is a formula (ready?)

Sigma =SQRT(LN(1/(1-EXP(-B24/1000000))^2))-(2.55155+0.802853*SQRT(LN(1/(1-EXP(-B24/1000000))^2))+0.010328*SQRT(LN(1/(1-EXP(-B24/1000000))^2))^2)/(1+1.432788*SQRT(LN(1/(1-EXP(-B24/1000000 ))^2))+0.189269*SQRT(LN(1/(1-EXP(-B24/1000000))^2))^2+0.001308*SQRT(LN(1/(1-EXP(-B24/1000000))^2))^3)+1.5

where B24 is DPMO

This formula is accurate to two decimal places between Sigma = 3.2 - 9.0, and one decimal between 2.5 & 3.2, and shouldn't be used for Sigma < 2.5

If you're using Excel you can use:
Sigma =NORMINV(1-B24/1000000,1.5,1)

The Excel function is accurate to 2 decimal places between Sigma = 1.4 - 5.7, and to 1 decimal place for Sigma = 5.7 - 6.1

Check out: http://www.pyzdek.com/Sigma-PPM-Percent%20Converter.htm

Been there, I don't have the complete MS 2000 suite to utilize the converter.


Thanks for the feedback.


Jim

DPMO to sigma is simple. Go to normal tables with your DPMO, get the z value and add 1.5. That is your sigma value.

Use the paste function in Excel for the most accurate normal tables.

GaryC

DPMO to sigma is simple. Go to normal tables with your DPMO, get the z value and add 1.5. That is your sigma value.

Use the paste function in Excel for the most accurate normal tables.

GaryC

In Excel try:
DPMO-->sigma -NORMSINV((DPMO)/1000000)
Sigma-->DPMO (1-NORMSDIST(Sigma))*1000000
Excel isn't accurate for DPMO's<.4 or sigma's over about 5.3.
This is of course without the 1.5 gimmie.