Tests of three fixed Kepler-13A periods in twelve TESS sectors depend on flux and noise assumptions

We test two reported periods near 25.4 hours and an exact 5:3 orbital frequency ratio in twelve TESS sectors from 2019 to 2024. We fit one coherent sine at each fixed period. Both flux products use the same samples, bins, and nuisance model. Under independent, symmetric sector noise scores, a complete sign test permits arbitrary time correlation within a sector. The main PDC fit gives 28.09 ppm at 25.426 hours with p=0.01074. SAP gives 24.74 ppm at 25.389 hours with p=0.01123. Neither passes a 5% correction for the six product and period tests. At 25.426 hours, the PDC p value changes to 0.08057 with linear trends and 0.002441 with 15-minute bins. The latter passes a six-test correction within that model. Paired difference tests do not establish a coherent difference between the products at 5%. Five observing-year sign groups have a minimum possible p value of 0.0625 and cannot reject at 5%. We derive and implement a conservative amplitude upper bound by test inversion. The bounds depend on the same sign assumptions and do not exclude the earlier reported 21 ppm signal. These tests do not establish a stellar rotation rate or show that an evolving light signal has stopped. The package preserves raw data, provenance, paired data, code, all fits, figures, a proof, and reproduction instructions.

Kepler-13A · TESS · stellar variability · fixed-period tests · 25.4-hour signal · sign symmetry · correlated noise · confidence bound · public data · reproducible photometry