DataPage+ and Profile

Got a question regarding Profile. I have the following data and I'm having trouble with what's being reported.

Data point 1: Profile MAX is 0.010 Profile MIN is -0.100 giving me a Profile callout of 0.110

Data Point 2: Profile MAX is 0.005 Profile MIN is -0.075 giving me a Profile callout of 0.080

Data Point 3: Profile MAX is -0.002 Profile MIN is -0.126 giving me a Profile callout of -0.126

I get that the first two are giving me the range, and the third one is giving me the WORST value, but shouldn't it be an absolute value like the first two?

I have 12 data points in this study and I have one part giving us heartburn because it doesn't report the absolute value.

TIA
John

Randy L. over 12 years ago

John,

If you are reporting proflie to a datum your information is reporting correct. You can report form only or form & location. If you want absolute values you would have to report as form only. Remember reporting form and location you are reporting worst case deviations positive and negative from zero. So in your case for Data point 3 you have no positive deviation, only negative and -.126 is the largest deviation from zero therefore you profile is -.126.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
John Kugler over 12 years ago

OK, then why is Data Point 1 0.110 not -.100. I understand what your saying, it just doesn't make sense for all the points.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
Randy L. over 12 years ago

Profile controlling "form and location" = Max positive deviation from zero + Max negative deviation from zero.
Profile controlling "form only"= absolute deviation.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
ir a programmer over 12 years ago

OK, then why is Data Point 1 0.110 not -.100. I understand what your saying, it just doesn't make sense for all the points.

The first two have a range of plus and minus material, making the total deviation the sum of the absolute values. The third example is all minus material - there is no sum of absolutes to calculate, making the total deviation a "worst case" to report.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
John Kugler over 12 years ago

Sorry, I may be thick headed, but if I change the MAX value of point 3 to a POSITIVE 0.002, I'll get a profile callout of 0.128. Which then becomes a swing of .254 from negative to positive.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
John Kugler over 12 years ago

The first two have a range of plus and minus material, making the total deviation the sum of the absolute values. The the third example is all minus material - there is no sum of absolutes to calculate, making the total deviation a "worst case" to report.

Yes, I understand that.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
John Kugler over 12 years ago

Here is my raw data. First column is the "Profile" callout. This is what PCDMIS reports. The middle column is the MAX value. The last column is the MIN value. See how the middle one is so different than the rest??

0.155 0.079 -0.076
0.131 0.031 -0.1
0.145 0.031 -0.113
0.15 0.031 -0.12
0.129 0.005 -0.124
0.158 0.021 -0.138
-0.126 -0.002 -0.126
0.151 0.025 -0.126
0.163 0.041 -0.121
0.161 0.041 -0.12
0.144 0.022 -0.123
0.131 0.009 -0.122
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
ir a programmer over 12 years ago

Without being familiar with the part, I would almost guess that middle one may have bad hits (only takes one bad one to wonk it up), or perhaps that feature is really "low" in comparison to the others?
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
Randy L. over 12 years ago

How do you get a swing of .254? Going from a -.002 to a +.002 = .004 swing which agrees with the profile callout of .128.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
John Kugler over 12 years ago

The reported profile callout (In PCDMIS) goes from -0.126 (With a MAX of -0.002 and a MIN of-0.126) to +0.128 (Max of +0.002 and a MIN of -0.126)

That's my swing and what is giving me the most heartburn.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel