What Capability Index should I use? SMT Application

Hello guys...

I would like what index is suited for SMT.

Ex. 0.1 +0.05/-0.2

Is it Ppk or Cpm?

Thank you

SAMURAI X said:

Hello guys...

I would like what index is suited for SMT.

Ex. 0.1 +0.05/-0.2

Is it Ppk or Cpm?

Thank you

Click to expand...

Hello Samurai X,

Welcome to The Cove.

I am having a problem understanding what you are asking. Would it be possible to give a little more information in your question?

Thanks, and again, welcome.

Dave

This depends on where your data comes from

PPK is for the population without sampling
Cpk is for spc

\The difference is where the values for the std deviation comes from.
For PPK it is from the numbers for CPk it is from the formula with d2 in the spc charts

SAMURAI X said:

Hello guys...

I would like what index is suited for SMT.

Ex. 0.1 +0.05/-0.2

Is it Ppk or Cpm?

Thank you

Click to expand...

Have you calculated your DPMO?

Are you familiar with IPC 7912?

michellemmm said:

Have you calculated your DPMO?

Are you familiar with IPC 7912?

Click to expand...

IPC 7912 information.

Stijloor.

michellemmm said:

Have you calculated your DPMO?

Are you familiar with IPC 7912?

Click to expand...

It's possible that introducing DPMO might overcomplicate things if the OP is just taking linear measurements of the location of a single component (assuming that the "SMT" referred to is Surface Mount Technology and we're talking about circuit boards). He might have some data that he wants to derive a capability index from, in which case knowing the number of "opportunities" on the whole assembly might be irrelevant. Just guessing, though.

Stijloor said:

IPC 7912 information.

Stijloor.

Click to expand...

Here is a guide for calculating DPMO.

All I can say is.....It is an eye opener!

Jim Wynne said:

It's possible that introducing DPMO might overcomplicate things if the OP is just taking linear measurements of the location of a single component (assuming that the "SMT" referred to is Surface Mount Technology and we're talking about circuit boards). He might have some data that he wants to derive a capability index from, in which case knowing the number of "opportunities" on the whole assembly might be irrelevant. Just guessing, though.

Click to expand...

Jim,
Five years ago, I would have agreed with you. One of our customers forced me to try DPMO before declaring it redundant. The first time I used it, I was amazed. It's like someone changing the FMEA RPNs. It will help calculating capabilities for the right parameter/reason.

Good day,

Sorry for my incomprehensive question.

As what I understood, for symmetrical tolerances (ex.0.2±0.05) I will going to use Cp&Cpk (s=Rbar/d2) or Pp&Ppk (s=actual sigma) and one sided
tolerances (ex. 0.08 max) its Cpu or Cpl and Ppu or Ppl.

What I am confusing about is for the unsymmetrical tolerance (Ex. 0.5 +0.3/-0.1). Should I use Cpm (sigma is base on target value)?? Or I can still use the
Cp&Cpk and Pp&Ppk for this type of tolerance?

Hoping to clear my mind on this one.

More power to all.

D.Scott said:

Hello Samurai X,

Welcome to The Cove.

I am having a problem understanding what you are asking. Would it be possible to give a little more information in your question?

Thanks, and again, welcome.

Dave

Click to expand...

SAMURAI X said:

What I am confusing about is for the unsymmetrical tolerance (Ex. 0.5 +0.3/-0.1). Should I use Cpm (sigma is base on target value)?? Or I can still use the
Cp&Cpk and Pp&Ppk for this type of tolerance?

Click to expand...

So long as the tolerance is bilateral you can use Cpk or Ppk.