Search the Elsmar Cove!
**Search ALL of Elsmar.com** with DuckDuckGo including content not in the forum - Search results with No ads.
  • Google has changed ad sizes for some reason. I am investigating and will get the sizes reduced to what they have been until now. I apologise for the inconvenience.

Can we use pin gauges to measure an accurate Cpk ?

#1
We manufacture printed circuit boards. we have a hole with a specification of 1.00 mm +0.09 mm - 0.00 mm with a Cpk of greater than 1.33.
can we use pin gauges in increments of 0.01 mm to measure repeatedly and calculate the Cpk to the above specification ?
 

Miner

Forum Moderator
Staff member
Admin
#2
It is theoretically possible, but would be difficult in reality. You would have to be on target with a standard deviation of approximately 0.01, including the measurement variation.
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#3
maybe a different approach is in order rather than a simple march to calculating a Cpk? I assume(?) that the characteristic is the hole diameter and not a positional characteristic? what does the hole do?
 

bobdoering

Stop X-bar/R Madness!!
Trusted
#6
We manufacture printed circuit boards. we have a hole with a specification of 1.00 mm +0.09 mm - 0.00 mm with a Cpk of greater than 1.33.
can we use pin gauges in increments of 0.01 mm to measure repeatedly and calculate the Cpk to the above specification ?
First, to meet 1.33, you would have to be running at 75% of your tolerance or better - meaning the bulk of your data is within .067mm spread. To get decent resolution to determine your distribution, you would need measurement of at least 10 "buckets" - or about 5 micron increments. If your distribution is skewed, you might need to do better than that. A "tenth" set of pins (.0001" or 2.5 micron) would give you some idea what your distribution looks like. Yes, pins are MMC plus fit - but when looking for features that have mating features inserted into them, it is a darn safe measurement technique. Most customers would agree to it. From my experience with circuit board production, I would be comfortable with it.
 
#7
Hi Ian,
I agree with Miner, the manufacturing process standard deviation would have to be just over 0.011 mm or better and your pin increments swallow nearly all this up. You could go for smaller pin increments but practically how much can you really "feel" when applying what are essentially plug gauges?
There's a difference between sentencing parts and measuring a feature to ascertain process capability, it may be easier to persuade the customer that go/no go checks are more appropriate for this product than finding a device to measure a hole feature of this size which behaves consistently.
 
#8
There is a possibility to calculate process capability with go/no-go checks using the Z-scores, if calculating process capability is a customer requirement.

Cpk = Z Score (p)/3
 

bobdoering

Stop X-bar/R Madness!!
Trusted
#9
Let's not forget, customers often rubber stamp 1.33 Cpk requirements on things with no regard to the practicality of the demand. There is a hole check device for PC holes - but I bet tenth set of pins would be more consistent. Try a gage R&R and see which is best.
 

Al Rosen

Staff member
Super Moderator
#10
Let's not forget, customers often rubber stamp 1.33 Cpk requirements on things with no regard to the practicality of the demand. There is a hole check device for PC holes - but I bet tenth set of pins would be more consistent. Try a gage R&R and see which is best.
This is the hole check device you are referring to.
 

Attachments

Top Bottom