For astronomical instruments, accurate knowledge of the optical pointing and coupling is essential to characterize the alignment and performance of (sub-)systems prior to integration and deployment. Ideally, this requires the phase response of the optical system, which for direct (phase insensitive) detectors was not previously accessible. Here, we show development of the phase-sensitive complex beam pattern technique using a dual optical source heterodyne technique for a large-field-of-view microwave kinetic inductance detector camera at 350 GHz. We show here how you can analyze the measured data with Fourier optics, which allows integration into a telescope model to calculate the on-sky beam pattern and telescope aperture efficiency prior to deployment at a telescope.

Complex field mapping is a powerful tool to characterize the optical performance of astronomical instruments, and has become the standard for characterizing heterodyne array cameras. Recently, an adaptation of the heterodyne beam mapping technique was demonstrated on a single pixel of a direct detector instrument. We present a novel measurement apparatus and data acquisition techniques to efficiently reconstruct the complex field pattern of individual pixels across a direct detector focal plane array. These techniques are scalable to high pixel counts as the technology maturation and scientific requirements push to larger arrays. For this demonstration, we used an engineering model of the low-frequency band of the APEX microwave kinetic inductance detector camera with a center frequency of ν = 350 GHz. Amplitude and phase radiation patterns were measured from all 880 pixels of the test array in two orthogonal polarizations. We also discuss an updated postprocessing pipeline using the complex field data to characterize the optical performance of the array. Using the measured complex field pattern, we extract the co- and cross-polarization patterns and Gaussian beam parameters, and propagate the beam from the measurement plane to additional planes of interest across all pixels in the test array. Complex field measurements of direct detectors allow more precise characterization of beam parameters when compared to thermal measurements, particularly for individualized fitting in postprocessing not reliant on the accuracy of the probe system alignment. These techniques enable high-precision characterization of individualized beam parameters as well as the overall optical system to very large format arrays with modest computational processing power. These results demonstrate the diagnostic power of the presented measurement and analysis techniques.

Large ultrasensitive detector arrays are needed for present and future observatories for far infrared, submillimeter wave (THz), and millimeter wave astronomy. With increasing array size, it is increasingly important to control stray radiation inside the detector chips themselves, the surface wave. We demonstrate this effect with focal plane arrays of 880 lens-antenna coupled microwave kinetic inductance detectors (MKIDs). Presented here are near field measurements of the MKID optical response versus the position on the array of a reimaged optical source. We demonstrate that the optical response of a detector in these arrays saturates off-pixel at the ∼-30-dB level compared to the peak pixel response. The result is that the power detected from a point source at the pixel position is at a similar level to the stray response integrated over the chip area. With such a contribution, it would be impossible to measure extended sources, while the point source sensitivity is degraded due to an increase of the stray loading. However, we show that by incorporating an on-chip stray light absorber, the surface wave contribution is reduced by a factor >10. With the on-chip stray light absorber, the point source response is close to simulations down to the ∼ -35-dB level, the simulation based on an ideal Gaussian illumination of the optics. In addition, as a crosscheck, we show that the extended source response of a single pixel in the array with the absorbing grid is in agreement with the integral of the point source measurements.