T CrB is one of the most-famous and brightest novae known, and is a recurrent nova with prior eruptions in 1866 and 1946 that peak at 𝑉=2.0. I have constructed light curves spanning 1842-2022 with 213,730 magnitudes, where the 𝐵 and 𝑉 magnitudes are fully corrected to the Johnson system. These light curves first reveal a unique complex high-state (with 20× higher accretion rate than the normal low-state) stretching from -10 to +9 years after eruption, punctuated with a deep pre-eruption dip (apparently from dust formation in a slow mass ejection) and a unique enigmatic secondary eruption (with 10 per cent of the energy of the primary eruption), with the light curves identical for the 1866 and 1946 eruptions. Starting in 2015, T CrB entered the high-state, like in 1936, so a third eruption in upcoming years has been widely anticipated. With the pre-1946 light curve as a template, I predict a date of 2025.5±1.3 for the upcoming eruption, with the primary uncertainty arising from a possible lengthening of the pre-eruption high-state. I use the large-amplitude ellipsoidal modulation to track the orbital phase of the binary from 1867-2022. I measure that the orbital period increased abruptly by +0.185±0.056 days across the 1946 eruption, the 1947-2022 years had a steady period decrease of (−8.9±1.6)×10 −6 days-per-day, and the 1867-1946 years had a steady period change consistent with zero, at (+1.75±4.5)×10 −6 days-per-day. These large period changes cannot be explained by any published mechanism.