An independent day model gives approximate limits below 0.15 electron/s for four beat harmonic differences in matched AR Scorpii TESS Sector 91 data with zero quality flags

We compare simultaneous 20-second and 120-second TESS light curves for AR Scorpii in Sector 91. Each slow sample is paired with the mean of the same six fast exposures. We retain 9,948 pairs with zero quality flags. A shared Fourier fit uses known spin and orbital clocks. The four beat harmonic coefficient differences have amplitudes from 0.0128 to 0.0418 electron/s. With 19 independent day groups and an approximate F calibration, their simultaneous 95% upper limits are 0.1245, 0.1377, 0.1348, and 0.1468 electron/s. Each limit is below 0.07% of the median aperture flux. We find no clear difference under this model. Relaxed quality cuts give weaker limits. The nominal linear 120-second response is only about 0.015 at these frequencies. Thus the observed limits give weak constraints on differences in the original pulses. The result applies to these selected public products and assumes independent day errors. We preserve both source files, hashes, code, paired residuals, numerical checks, controls, three figures, and reproduction instructions.

TESS · AR Scorpii · white dwarf pulsars · exposure integration · paired photometry · correlated errors