RF Impedance-Matching Network

Automatic, source, and bias RF matching networks each tune impedance for a specific part of a dual-generator plasma system, and the "automatic" part matters because a plasma's impedance isn't fixed — it shifts constantly as the process runs, requiring continuous retuning just to keep power delivery efficient.

Automatic tuning exists because the plasma's impedance is a moving target throughout a process step, not a fixed value the match network can be set once and left alone. As gas chemistry shifts, pressure changes, or a film deposits on chamber surfaces during a run, the plasma's electrical impedance drifts along with it; an automatic matching network continuously senses reflected power and adjusts its internal variable capacitors in real time to keep tracking that moving impedance, maintaining efficient power delivery throughout the process rather than only at the single condition a static, manually-tuned match was set for.

Source and bias matching have to be handled as two entirely separate tuning problems, because the source and bias generators see different load impedances from different parts of the plasma system. The source generator's matching network tunes against whatever impedance the plasma coil or source electrode presents; the bias generator's matching network separately tunes against the chuck's own impedance as seen through the plasma sheath — these are genuinely different electrical environments, so sharing one matching network between the two generators isn't an option, and each needs its own dedicated, independently auto-tuned match.

Automatic, source, and bias RF matching networks together solve the continuous impedance-tracking problem that a dual-generator plasma system creates, and getting both matches right independently and in real time is what actually determines whether the source and bias power levels a recipe specifies end up delivered to the plasma as intended.