TESS and ZTF data exclude a stable 235.587-day repeat of the K2 eclipse of EPIC 220208795 in September 2025

EPIC 220208795 showed a single deep K2 eclipse in 2016. Earlier work proposed candidate periods of 235.587 and 290.230 days from sparse ground data. We test a stable repeat of the original eclipse with public TESS sector 97 pixels and ZTF DR24 photometry. At the 235.587-day prediction on 2025-09-24, TESS has 73 cadences in the part where the K2 template predicts more than 10% flux loss. No deep eclipse occurs. The fitted depth multiplier is -0.00766 with conditional standard error 0.01783, where 1 repeats the 28.33% K2 template. Three apertures give consistent results. All 42 control windows recover quarter-depth injected eclipses. The result holds on a 201-point period grid from 235.487 to 235.687 days. A separate two-hour fit with a 3% error floor also excludes a full repeat. An independent ZTF red point shows a magnitude loss of 0.00513 plus or minus 0.02644, where the fixed template predicts 0.14294. These data exclude a stable 235.587-day repeat at this epoch under the stated shape and cross-filter depth assumptions. They do not exclude a changing disk or other periods. The saved ZTF data do not usefully sample the 290.230-day eclipse. We supply source data, exact requests and hashes, code, figures, results, and reproduction instructions.

EPIC 220208795 · TIC 336889445 · TESS · K2 · ZTF DR24 · eclipses · circumsecondary disk · repeat test · archival photometry