I work for an RF Amplifier Manufacturer. We have three test benches that are used to test product. Basically, the unit to be tested is placed in a fixture and then a button is pressed to start the test. The computer determines if the unit passes or fails. The operator is not required to do anything further to determine if the unit passes or fails. My question is how should I set-up the GR&R portion of my MSA?:
1) Classify operators as appraisers and have them test several units on the same bench.
2) Classify the test benches as appraisers (since the operator's job of simply pushing a button should not affect reproducibility) and test several units on each of the benches.
Send to friendSign In to get started!
Comments
qdigest 2/25/2003
Depending on the criticality of the part/feature you're measuring, 26% would be considered acceptable. What does your customer who requested the study require for %R&R?
As for the difference between you and the customer:
1) How much of the %R&R is due to appraiser variation and equipment? If the contribution of appraiser variation is very high as compared to equipment, the opportunity to determine the source of between appraiser differences and correct them can reduce the total %R&R significantly. It may be more difficult to reduce the equipment variation.
2) Did the customer use a different type of equipment or just a different instrument of the same type that you used?
3) Was there any difference in measuring technique between you and the customer?
4) Different groups of parts may contribute more variation to the study. Preferrably, both parties would measure the same group of parts withthe same instrument.
Hope this helps. It's hard to be more specific without more detail on what you're measuring and with what type of instrument.