Verify coherent gain in one output

Define the claim before choosing a detector

A useful experiment asks whether changing relative phase directs more of a fixed supplied power into a specified output mode. That question is narrower and more testable than whether combining makes light stronger. Define the target by an output port, a calibrated spatial aperture, or coupling into a known mode. Define the input reference plane and the power included in the denominator.

For a two-input demonstration, an ideal balanced combiner gives a simple prediction. For a tiled array, the target may be a main-lobe angular region with sidelobes measured separately. These setups require different collection procedures. The experiment should be designed around the claimed metric, because a detector that captures an arbitrary fraction of the field cannot establish a universal combining efficiency.

Calibrate the detector and collection path

Check detector linearity over the full expected signal range using known attenuation or another suitable calibration. A saturated detector can flatten maxima and distort visibility, while automatic gain control can create apparent changes unrelated to optical power. Record background with the inputs blocked and quantify its variation. Keep detector settings fixed during comparisons unless a calibrated conversion is applied.

Measure the transmission of any lenses, filters, fibers or apertures between the combiner and detector. If the target is a fiber mode, coupling depends on position, angle and wavefront as well as total incident power. If the target is a camera region, state its boundaries and pixel calibration. These details determine whether the reported increase concerns total output, spatial concentration, or improved coupling into the chosen measurement mode.

Measure each input independently

Record the power of each input at the same reference plane used in the efficiency definition. Use taps or calibrated interleaved measurements that do not change source operation unnecessarily. Blocking one arm downstream is often preferable to switching a laser off and on, because source startup can change frequency, pointing or power. Keep a record of the individual signals throughout the combined measurement.

For a balanced two-port device with one input of power P and the other blocked, each ideal output receives P/2. This baseline is crucial: comparing the two-input bright output with only one of those half-power outputs gives a factor of four, while the total supplied power has doubled. The experiment should show both ratios so that a local enhancement is not mistaken for net energy amplification.

Scan phase through a full cycle

Let the inputs have powers P₁ and P₂. For the ideal balanced combiner, P₊(φ)=[P₁+P₂+2√(P₁P₂)cosφ]/2 and P₋(φ)=[P₁+P₂−2√(P₁P₂)cosφ]/2. Scan the phase through a full cycle and fit both outputs simultaneously after detector calibration. The predicted sum is P₁+P₂, independent of phase.

With P₁=P₂=1 watt, the desired output ranges from zero to two watts while the other changes complementarily. With P₁=4 watts and P₂=1 watt, the range is 0.5 to 4.5 watts. These calculated examples test amplitude balance as well as phase. A lower maximum can result from mode mismatch or losses, so compare it with independently measured overlap and transmission rather than assigning every deficit to phase noise.

Keep input power fixed during the phase change

A phase actuator can inadvertently change amplitude, polarization or pointing. Heating a path changes refractive index but may also move the beam; changing laser current can alter both optical frequency and output power. Measure these cross-couplings before interpreting the target increase. If input powers change, include that change in the normalized efficiency rather than attributing all output variation to coherent addition.

An independent phase reference or a calibrated actuator response helps establish the horizontal axis of the scan. Repeat the scan in both directions to reveal drift or hysteresis. Interleave a fixed reference setting when the experiment lasts long enough for source changes to matter. These controls make the phase dependence a reproducible physical relation instead of a correlation between two slowly changing instrument readings.

Measure where the remaining power goes

For a discrete-port combiner, monitor all significant outputs. For a phased array, map enough of the angular pattern to quantify sidelobes and power outside the target region. For a lossy optical path, measure or bound absorption and scattering losses. A target-mode efficiency alone is useful, but a complete power budget explains whether a deficit represents unwanted routing, collection loss or dissipation.

During transients, include stored-energy changes if the setup contains a cavity or active source feedback. A temporary output rise can draw on stored energy without changing the steady input immediately. The steady balance is simpler only after the relevant transients settle. State the observation window and whether the claimed efficiency is instantaneous, pulse-integrated or a long-time average.

Use a defined incoherent or randomized baseline

A useful comparison randomizes relative phases while keeping input powers and geometry fixed. In a symmetric N-input ideal combiner with equal powers and independent uniform phases, the mean target fraction is 1/N. Perfect alignment raises it to one. The enhancement relative to that baseline is N, while the target output still cannot exceed the total passive input power.

A finite number of random settings has statistical variation, and a physical phase randomizer may also change other properties. Record the actual phase sampling or use enough independently verified settings to estimate its mean and uncertainty. Do not choose an unusually dim random realization as the baseline and present the resulting ratio as a typical enhancement. The comparison distribution is part of the result.

For four equal inputs, the ideal random-phase mean target fraction is 0.25, whereas the aligned fraction is one. Comparing one aligned realization with one random realization is unreliable because the random target intensity fluctuates. Average over a stated ensemble and report its spread. If phase settings are correlated, the effective sample count is smaller than the number of recorded frames.

A useful cross-check is to predict the target signal from independently measured input phases and amplitudes, then compare that prediction with a detector record not used to fit those quantities. Agreement supports the coherent-sum mechanism directly. Disagreement can reveal unmonitored modes, phase-reference errors or detector calibration problems. This independent prediction is stronger evidence than an optimizer simply finding a setting that makes its own feedback detector brighter.

Test stabilization under a controlled disturbance

An active combining system should be tested after a known phase disturbance or under a specified drift. Record the target power, input powers and actuator commands over time. The recovery time and residual fluctuations reveal what the controller actually maintains. MIT Lincoln Laboratory's described architecture uses output monitoring for phase regulation, motivating this type of closed-loop characterization without supplying a universal performance guarantee.

Compare controller-on and controller-off operation under comparable conditions. If the controller also changes source power, report that effect. A stable mean can hide rapid fluctuations that matter for a target application, so retain time traces and spectra where appropriate. A single optimized snapshot establishes alignment at one moment; sustained combining requires evidence over the time scale for which stability is claimed.

Report a recovery criterion, such as returning within a stated fraction of the pre-disturbance efficiency, rather than choosing a favorable point by eye. Also state the disturbance amplitude and whether the actuator saturated. A controller that recovers quickly from a small phase step may behave differently after a full-cycle excursion or a simultaneous amplitude change. The transient test should match the operating conditions for which reliability is claimed.

Quantify uncertainty in the efficiency

Define η=Ptarget/ΣPinput using calibrated quantities at the stated planes. Propagate uncertainty in detector responsivity, tap ratios, background and collection fraction. Correlated calibration errors can cancel or reinforce, so do not assume every uncertainty is independent. If the target power and input sum share a common monitor, explain how that common factor enters the ratio.

Repeat the measurement after realignment or on separate runs to assess reproducibility. Report the spread and any systematic limitations, not just the best observed value. An efficiency slightly above one in a nominally passive setup usually calls for checking calibration, omitted inputs or transient storage before making a new physical claim. The normalized lossless model provides a strong consistency test precisely because its upper bound is explicit.

State the achieved result concretely

A strong result reports that a measured fraction of supplied power reached a defined mode under specified phase control, with uncertainty and stability limits. It shows the phase scan, input monitoring and complementary-channel behavior that support the interpretation. If peak intensity rather than mode power is the application metric, include the spatial profile and pulse duration needed to interpret that peak.

The experiment then demonstrates coherent routing or concentration with a clear energy account. It does not need an exaggerated claim of energy multiplication to be useful. The practical achievement is that phase control makes more of the available power useful at the target, and the measurements identify how much improvement came from that control rather than from a change in source power or collection geometry.

Go to the original work.

These links identify the sources behind specific claims. Verification scope varies by source and is stated below.

1. 1 / University teachingMIT: Mach–Zehnder interferometers and beam splitters ↗

   Beam splitters combine quantum path amplitudes.

   Verification scope: Official course record inspected; the matrices and worked examples here are independently specified.

2. 2 / Official researchMIT Lincoln Laboratory: Coherent Beam Combination in a Laser Oscillator Using Active Phase Control ↗

   A described beam-combining architecture uses a photodetector and feedback to regulate phases of multiple laser elements.

   Verification scope: Official technology description inspected; no advertised efficiency or scalability claim is treated as a universal bound.