You are talking about attribute data, not a variable. A variable is a
measurable item: the thickness of a board, a chemical concentration, the
diameter of a screw. This type of data typically follows a normal
distribution about a mean. Attribute data (yields) is non-normal since you
can't have a defect level less than 0. Depending on what you are measuring,
you're distribution is going to be Weibull, Poisson or binomial.
You can do some mathematical and statistical juggling to take a number like
800 ppm and back out a process capability (Cp, Cpk) but I'm pretty sure the
values have no statistical validity. It's been a while since I've done
this, but the statistic tools package in Excel can do most of this stuff.
You'll need a statistical quality control textbook though to decipher what
the different variables are.
For what it's worth, I'm personally opposed to using SPC in this manner. I
think too much emphasis is put on the numerical value of the capability
rather than the process of using SPC as a tool to improve quality. I've
seen too many instances where gathering and analyzing data are
short-circuited or quickly cast aside after some magic process capability
value was attained. I recall an occasion where I was working with some
yield data that was doing well, so the control limits were tightened to
flush out another source of variation. Of course we started seeing out of
control warnings. After a few days, a manager button-holed me and wanted to
know why my process was out of control now and when I was going to get the
problem fixed. I tried explaining what was happening and that we should
never stop getting out of control warnings, but didn't make much progress.
(Get down off soapbox).
Mike McMaster
RF Product Engineer
Merix Corporation
503-992-4263
----------
From: Collins, Graham [SMTP:[log in to unmask]]
Sent: Tuesday, February 08, 2000 9:46 AM
To: [log in to unmask]
Subject: Re: [TN] Subj: Six Sigma
Is there a formula one can use to come up with these numbers? E.G.
if my
process yield last month was 800 PPM, can I crunch that number and
say we
were at 4.8 sigma?
- Graham Collins
-----Original Message-----
From: Hollandsworth, Ron [mailto:[log in to unmask]]
Sent: Tuesday, February 08, 2000 11:29 AM
To: [log in to unmask]
Subject: Re: [TN] Subj: Six Sigma
Dave:
Thanks this is what I was looking for. I will chart this out and
pass it on
to the Mfg. Engineer who ask the question.
Thanks to all the replys. (Terri, Lou, & Ed)
Appreciate the quick response.
Ron Hollandsworth
-----Original Message-----
From: David Fish [mailto:[log in to unmask]]
Sent: Tuesday, February 08, 2000 1:09 PM
To: Hollandsworth, Ron
Subject: Re: [TN] Subj: Six Sigma
Ron: From Breyfogle (ISBN 0471296597):
Spec limit|Defective ppm - no mean shift|Defective ppm - 1.5 sigma
mean
shift
±1 sigma|317300|697700
±2 sigma|45500|308700
±3 sigma|2700|66810
±4 sigma|63|6210
±5 sigma|0.57|233
±6 sigma|0.002|3.4
Dave Fish
"Hollandsworth, Ron" wrote:
> Good Morning TechNetters!
> Here is a brain teaser for this early morning. Thinking six
sigma. Six
> sigma is suppose to be 3.4 PPM. Good! Now what is five sigma,
four,
three,
> two , one. There is a chart that shows that relationship of a
certain
sigma
> and the PPM associated with that particular sigma value.
>
> Ok, the first five to get the right answer will be eligible to go
on to
the
> "Who wants to be a Black Belt" show. Enough levity, can someone
help me
> out. Do you know the chart I am referring to? Can you send me
the values
> and I'll make a chart.
>
> Grasshopper Ron
>
> ##############################################################
> TechNet Mail List provided as a free service by IPC using LISTSERV
1.8c
> ##############################################################
> To subscribe/unsubscribe, send a message to [log in to unmask] with
following text in
> the body:
> To subscribe: SUBSCRIBE TECHNET <your full name>
> To unsubscribe: SIGNOFF TECHNET
> ##############################################################
> Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
additional
> information.
> If you need assistance - contact Gayatri Sardeshpande at
[log in to unmask] or
> 847-509-9700 ext.5365
> ##############################################################
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV
1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with
following
text in
the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
additional
information.
If you need assistance - contact Gayatri Sardeshpande at
[log in to unmask] or
847-509-9700 ext.5365
##############################################################
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV
1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with
following text in
the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
additional
information.
If you need assistance - contact Gayatri Sardeshpande at
[log in to unmask] or
847-509-9700 ext.5365
##############################################################
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in
the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################
|
