Nine TESS PDCSAP transit times of HD 21749 b fit a constant period from 2018 to 2025 under a stated noise model

We measure nine transit times of HD 21749 b with public TESS data from 2018 to 2025. A constant period of 35.61343715 days gives chi-square 4.68 for seven degrees of freedom. The conditional period error is 1.59 seconds. This result uses corrected PDCSAP flux and local quadratic baselines. We use the largest timing error from profile fits and three residual block tests. A quadratic clock term improves chi-square by only 0.14. The latest transit is 96.15 minutes later than the discovery point prediction. The discovery period error alone grows to about 65 minutes there. This offset does not establish orbital change or an error in the source clock. Gan et al. (2021) and the saved ExoFOP record already give longer periods. We then fit both the transit shape and clock to only the first eight events. The last event arrives 1.10 minutes before that prediction, with a combined conditional error of 2.52 minutes. Tests with corrected flux, quadratic and cubic baselines, and three physical shapes give similar periods. Raw aperture flux and linear baselines produce excess scatter under the primary errors. Across all checks, the periods span 3.96 seconds. Thus, the errors do not cover all flux model effects. A separate bootstrap refits the shared shape and gives a 1.58-second period spread. We supply the times, event decisions, original light curves, source records, code, bootstrap draws, figures, and offline repeat instructions. The contribution is an independent timing dataset and consistency check. It does not claim a new planet or the first period update. These data do not exclude small timing variations.

HD 21749 b · GJ 143 b · TESS · transit timing · ephemeris · public data · model dependence · reproducible astronomy