Three quick calculators from the floor — process capability, yield & process sigma, and Gage R&R. Everything runs in your browser; nothing is uploaded. Each is pre-loaded with example data so you can see it work, then paste your own.
Results
Cp / Cpk use the within-subgroup (short-term) sigma estimated from the average moving range (σ̂ = MR̄ / 1.128). Pp / Ppk use the overall (long-term) sample sigma. DPMO and yield are estimated from the overall fit assuming a normal distribution.
Defects & sigma
Rolled throughput yield
Throughput yield uses the Poisson model Y = e−DPU. Process sigma is the short-term sigma level (Zbench + 1.5 shift), the common Six Sigma convention.
Variance components (ANOVA)
Gage R&R summary
Computed with the ANOVA method (two-way with operator×part interaction). If the interaction is not significant it is pooled into repeatability. Study variation uses 6σ; %contributions are reported on the standard-deviation (study-variation) basis. ndc = 1.41 × (PV / GRR). AIAG guidance: %GRR < 10% acceptable, 10–30% marginal, > 30% unacceptable; ndc ≥ 5.
Control limits
X̄–R uses the standard Shewhart constants (A₂, D₃, D₄, d₂) for the detected subgroup size (n = 2–10). I–MR uses a moving range of two (E₂ = 2.66, D₄ = 3.267). σ̂ is estimated from R̄/d₂ (or MR̄/1.128). These are the process's own control limits, not spec limits.
Required n — mean
Required n — proportion
Large-sample normal approximation. For a proportion with a small lot, enter the population to apply the finite-population correction.
Conversion
Reference table
Short-term sigma level = 3·Cpk + 1.5 (the +1.5 long-term shift convention). DPMO is the two-sided nonconforming fraction for a centered normal process, 2·Φ(−3·Cpk)·10⁶.