LIF Conversion Stage

Overview

A LIF Conversion Stage represents one physical optical element — a doubling crystal, a mixing crystal — in the LIF frequency-conversion chain between the tunable laser and the sample. Each configured stage appears as a row in the Conversion tab’s table (see LIF Frequency Conversion), where its inputs are wired and, for at most one stage, its output is marked as the excitation beam. The device itself only moves to and reports a local input-beam position (in cm⁻¹); the chain’s topology — which stage feeds which, and which stage is the excitation beam — is per-experiment state configured on the Conversion tab, not part of the device profile.

Settings

Most LIF Conversion Stage settings are exposed in the hardware dialog with inline labels and tooltips, so they need no additional explanation here. A few items are worth highlighting:

  • Conversion Operation sets the stage’s operation — NHG (N-th harmonic), SFG, or DFG. Concrete drivers usually pin this to a fixed value that matches the physical device (a doubler is always NHG, for example), so it is not normally something you choose yourself.

  • Harmonic Order sets the harmonic order N for an NHG stage; it is ignored for SFG/DFG stages. Some drivers fix this at profile creation rather than leaving it freely editable (see Sirah FCU below).

  • Verify Move controls whether a move is confirmed against the stage’s read-back position after every move. When enabled (the default), a mismatch beyond Verify Tolerance fails the move; when disabled, a mismatch is only logged as a warning and the move is treated as successful.

  • Verify Tolerance (cm-1) sets the read-back window, in cm⁻¹, used by Verify Move. Defaults to 1.0.

Drivers

Virtual

A dummy driver.

Fixed

Represents a conversion-topology node that is not under Blackchirp’s control — a crystal or compensator tuned by hand, or one you simply want represented in the chain’s math without automating its motion. Moves always report success at the commanded position, so the stage participates fully in the chain with no separate place to enter its state.

Sirah FCU

The Sirah Frequency Conversion Unit is a doubling-stage driver for a Sirah frequency-conversion unit — a separate Sirah instrument from the Sirah Cobra grating controller, connected on its own serial port with its own communication settings. Harmonic order defaults to 2, the common lone-doubler case; the default is set prominently at profile creation, but it remains changeable afterward through the Conversion tab’s Change harmonic… action (see Changing the Harmonic Order). The doubling crystal’s angle-to-wavelength tuning curve is evaluated through a selectable calibration scheme rather than a fixed formula; see that section for the schemes, their settings, and the offline calibration workflow that produces the numbers you enter here.

Sirah FCU Calibration

The Sirah FCU’s doubling crystal is angle-tuned: the fundamental wavelength that phase-matches at a given crystal angle is set by the crystal’s own dispersion, not by a diffraction law. The driver evaluates this fundamental-wavelength-to-motor-position tuning curve through a user-selectable Calibration Scheme rather than a single built-in formula. Blackchirp only evaluates a scheme’s already-fitted curve — it does not fit one itself; the fit is produced offline against a bench-recorded calibration run, as described in Offline Calibration Workflow below. Since the compensator is geared directly to the crystal stage, one calibration curve covers the whole device.

Calibration Scheme selects between the three schemes below. Changing it immediately shows only the fields relevant to the newly selected scheme in the hardware dialog and hides the rest — including the Crystal Stage Geometry table’s motor-motion fields (stageStartFreqHz, stageHighFreqHz, stageRampLength, stageMaxPos, stageBacklashSteps), which are only reachable from the dialog while Physical is selected. Hiding is a dialog display convenience, not a reset: a field’s stored value stays in effect even while it is not shown.

Physical

A best-effort Type-I second-harmonic phase-matching model for a BBO or KDP doubling crystal, combined with the sine-bar drive mechanics that turn a phase-match angle into a motor position. It needs only a handful of fitted numbers, extrapolates gracefully beyond the measured wavelength range because it follows the crystal’s actual dispersion, and is the best choice unless the crystal is tuned close to a 90° phase-match angle, where the physical curve turns very steep and a coarse fit through it is unforgiving.

Its settings, on the Crystal Stage Geometry table and alongside Calibration Scheme:

  • Crystal Type — BBO or KDP; selects which crystal’s dispersion the model uses.

  • Invert Phase Match — selects the alternate (−) branch of the phase-match relation; toggle this if a fitted curve runs the wrong direction.

  • stageCutAngleDeg and stageTemperature — the crystal’s optic-axis cut angle and an arbitrary-units dispersion fit knob (not a controlled physical temperature).

  • stageLeverLengthMm, stageLinearOffsetMm, stageAngleOffsetDeg, stageScrewPitchmmPerRev, and stageMotorResolutionStepsPerRev — the sine-bar drive’s mechanical constants, the same geometry documented for the Sirah Cobra grating stage.

These are the same numbers that fcu_fit.py prints after fitting a bench calibration run; type them directly into the corresponding fields rather than importing a CSV. By default the script fits only three of them — temperature, linear offset, and angle offset — and holds the cut angle and screw pitch fixed at the values you supply, fitting those two only when --fit-cut-angle or --fit-screw-pitch is passed.

Polynomial

Imported forward (wavelength → position) and inverse (position → wavelength) coefficient lists, one row per polynomial order. It makes no assumption about the underlying physics, so it fits whatever curve shape the data shows — but, unlike Physical, it only interpolates the fitted curve and needs denser, well-spread measurements across the whole tuning range to be reliable. A low-order polynomial is also a poor proxy for a steep near-90° curve.

Its only setting is the Polynomial Coefficients table (columns order / forward / inverse), populated by importing the CSV fcu_fit.py exports (see below) rather than typed in by hand.

Spline

An imported table of (wavelength, position) points; Blackchirp builds a monotone interpolating spline through them in each direction. Like Polynomial, it is model-agnostic and needs denser measurements than Physical, but a spline tracks a steep near-90° curve locally without the ringing a high-order polynomial fit is prone to, making it the more robust choice there.

Its only setting is the Spline Points table (columns wavelengthNm / positionSteps), also populated by CSV import.

Offline Calibration Workflow

Fitting parameters or curves for any of the three schemes happens outside Blackchirp, using the standalone scripts under python/tools/ in the Blackchirp source repository on GitHub. These are a specialty offline utility: they are not bundled with the Blackchirp application download and not part of the installable blackchirp Python package — obtain them by cloning or downloading the repository. See that directory’s README.md for the full command-line reference:

  1. Record a calibration run with fcu_measure.py. At each fundamental wavelength, hand-tune the doubling crystal at the bench (peak the doubled output, following the unit’s own calibration procedure), then let the script query the FCU’s motor position over its serial port and append a wavelengthNm;positionSteps row to a measurements CSV:

    python fcu_measure.py --port /dev/ttyUSB1 --output measurements.csv

    A --simulate mode records a synthetic curve instead, so the tool is exercisable without hardware. Hardware mode needs the pyserial package.

  2. Fit and visualize the run with fcu_fit.py, which fits all three schemes from the same measurements CSV in one pass and plots the resulting tuning curves and residuals (needs numpy, scipy, and matplotlib):

    python fcu_fit.py measurements.csv --crystal bbo --cut-angle 32.3 \
    --save calibration.png

    It prints the fitted Physical scheme’s parameters to the console (fitting temperature, linear offset, and angle offset by default, with cut angle and screw pitch held fixed unless --fit-cut-angle or --fit-screw-pitch is passed), and writes an order;forward;inverse CSV for the Polynomial scheme and a wavelengthNm;positionSteps CSV for the Spline scheme alongside the measurements file.

  3. Bring the result into Blackchirp. Open the Sirah FCU’s hardware dialog, select the matching Calibration Scheme, and either type the printed Physical parameters directly into their fields, or open the Polynomial Coefficients / Spline Points row’s Edit… button and, inside that sub-dialog, click Import CSV… to load the corresponding exported file (see Settings Tab for how array settings and CSV import work generally). The importer matches header columns to the table’s fields by name, so either exported CSV loads without further preparation.

Note

The Physical scheme is a best-effort BBO/KDP phase-matching model fit to your measured data — its functional form follows real crystal dispersion, but fcu_fit.py fits it independently for each unit and is not guaranteed to reproduce any vendor-internal calibration curve exactly. Record enough calibration points across your working range, and check the plotted residuals, before trusting a fit (Physical, Polynomial, or Spline) for production tuning.