how do you separate chamber condition from recipe when film is wrong

When a film comes out wrong, you must separate TWO causes: CHAMBER CONDITION (hardware/state: wall buildup, target wear, temperature drift, a bad part) vs RECIPE (the process settings: gas flows, power, temperature setpoints). HOW TO SEPARATE: 1) Check the hardware first — is there buildup, a worn target, a temperature anomaly, a recent PM/clean? 2) Run a KNOWN-GOOD reference/qualification — does a wafer that used to pass now fail? If the same recipe now fails, it's likely CHAMBER (condition changed), not the recipe. 3) Look at the TIMING — did it change after a clean, a PM, or a part change? 4) Compare to the recipe baseline — did anyone change a setting? The one-line frame: 'I separate them by changing one thing: if a recipe that used to pass now fails, the chamber changed — the recipe is the constant, so the hardware is the suspect.' Clean, systematic, and exactly the GPS diagnostic mindset (Lessons 41, 43).