Monochromatic Monitoring Simulator

The Mono Simulator is a 7-page wizard that simulates depositing your design under an in-chamber single-wavelength optical monitor. It is the monochromatic counterpart of the Broadband Monitoring Simulator and shares the same setup and playback; the difference is that each layer is cut from one monitoring wavelength using one of three classic termination strategies.

You configure the deposition conditions and the per-layer monitoring plan on the first five pages, run a single manufacturing experiment on page 6, and read the resulting performance on page 7.

Settings

Page 1: Deposition Rates. Per material, set the mean rate (nm/s), the RMS fluctuation, and the correlation time governing how slowly the rate drifts. The preview shows a sample rate-vs-time trace; Randomize draws a new one.

Page 2: Parameters Deviation. Per material, add systematic and random index shifts and a systematic inhomogeneity. The lower table excludes chosen layers from monitoring (cut on time) with an optional extra relative thickness error, and shutter delay (mean and RMS) models the cut lag.

Page 3: Monitoring System. Set the measured quantity and polarization, the angle of incidence, the scan interval, and the number of confirm scans a cut needs before it is accepted. Chip glass is the witness chip the monitor watches: it opens on the design substrate, and picking another material moves the monitor signal onto that glass, for a witness that is not the same glass as the part. The preview plots the ideal signal versus deposited thickness for the selected layer, with the cut point marked.

Page 4: Monitoring Wavelengths. The per-layer plan: for each layer choose the monitoring wavelength and the termination strategy:

• Turning point: cut when the monitor signal reaches an expected extremum. The order column picks which extremum.
• Level: cut when the signal crosses a target level in the expected direction.
• By time: cut after a precomputed time, with no signal feedback.

Auto λ (all) picks, for every layer, the wavelength and strategy whose cut is most precise. Quarter-wave layers keep the reference wavelength’s turning point, and its self-compensation, while the reversal is still detectable; once a stopband saturates the signal, its layers are monitored outside the band, and a layer no wavelength can serve is cut by time. Set all puts one wavelength on the whole run, and the strategy dropdown beside it applies one termination rule to every layer at once. Clicking a table row shows that layer’s ideal signal in the preview beside the table.

Page 5: Signal Errors. Add random noise (percent of the reading), the monitor’s absolute noise floor (percent of full scale, which does not shrink with the signal), and a slow drift to the single-wavelength signal; the preview shows the noisy signal for the selected layer.

Page 6: Deposition Simulation. Press Start to run one manufacturing experiment, then play it back on the interactive timeline. The bar chart shows the estimated / actual / target thickness of the current layer; the spectrum shows the theoretical guide curves against the as-built curve.

Page 7: Resulting Performance. Tabs show the manufactured vs. theoretical spectrum, per-layer relative and absolute thickness-error bars, and tables of as-built thicknesses and refractive indices.

The deposited side and the way the resulting spectrum is scored follow the surface mode set in the Design Editor, shown as a badge on the window. The monitor signal is read through the whole witness chip: the growing coating on its front face and its bare back face, added incoherently, the way a transmittance monitor actually sees a chip in the chamber. A bare chip of n = 1.52 glass reads 91.8 %, not the 95.7 % of the coated surface alone. The chip hangs in the chamber, so the signal is read with air above the growing coating whatever medium the design is embedded in.

How to read it

Match the strategy to the layer. When a layer’s thickness is close to a quarter-wave multiple at the monitoring wavelength, a turning point is precise and direction-blind, so it is the natural choice. For other thicknesses, pick a level cut at a wavelength where the signal slope is steep through the cut point, which gives the best precision. On page 7, layers with large thickness errors are the ones whose monitoring wavelength or strategy isn’t serving them; try the auto-wavelength suggestion or a different strategy, then re-run. Run the experiment several times to see the spread rather than a single outcome.

References

• H. A. Macleod, Thin-Film Optical Filters, 5th ed., Ch. 12.
• Tikhonravov & Trubetskov, Appl. Opt. 44, 6877 (2005).