In Reply to Parent Post by Tim Folkerts
I think I "reverse engineered" your numbers.
I bet you pressure gauge is actually calibrated in "bars": 1 bar = 10^5 Pa = just under 1 Atm. The resolution is 1 bar. The result, however, is displayed in PSI.
45 bars => 652 PSI
46 bars => 667 PSI
At least, it sound logical and it is something to pass by the engineers who know the equipment.
It sounds like you may need a pressure gauge with better resolution.
If this is the case, the question about SPC is meaningless until you have a gauge with adequate resolution. Once the gauge is acceptable, then you can plan the rational subgrouping for SPC.
Do not plan the subgrouping based on convenience, but on what information you want from the SPC. What variation do you want WITHIN the subgroup, and what variation do you want BETWEEN subgroups. Normally, you do not want time based variation within the subgroup, which is why you want consecutive samples. If you are forced to introduce time, or have an autocorrelated process (changes occur very slowly over time) such as extrusion, you would go with an IMR chart.