The ES Ceti light signal shifts 136 seconds ahead of the 2018 clock between the 2020 and 2025 TESS visits

We test the quadratic light clock in de Miguel et al. (2018) with public TESS data for the helium-transfer binary ES Ceti. A Fourier fit to 120-second light curves from sectors 30 and 97 gives a first-harmonic phase change of -136.5 +/- 8.5 seconds from 2020 to 2025, relative to that fixed clock. Negative values are early. The error is from independent day blocks and excludes the historical clock error. A day-block bootstrap gives a 95 percent interval of -152.5 to -119.3 seconds. The second harmonic gives -132.2 +/- 8.5 seconds. Raw flux, small target apertures, and pointing controls also show an advance. An exploratory 2026 core-pixel check gives -136.7 +/- 16.6 seconds relative to 2020. The measured phase change is modulo one 620.21-second cycle; we report the branch nearest the forecast. The published coefficient covariance is absent. Its marginal errors allow 24.8 to 63.4 seconds of standard error in the forecast difference, so the TESS measurement error alone cannot set the significance of a clock failure. Disk emission can also shift a light signal. These data provide later phase measurements and a reproducible forecast test. They do not establish a new orbital acceleration or its cause. The package includes source data, pixel checks, code, figures, provenance, and an offline reproduction route.

ES Ceti · TESS · AM CVn stars · orbital timing · ephemeris forecast · public data · phase measurement