Radio pulses confirm a fast-spinning redback binary with a changing companion

A suspected binary reveals its pulses

Astronomers have detected radio pulsations from PSR J2053+6905, confirming that a previously identified candidate is a redback pulsar binary. The neutron star rotates once every 1.68 milliseconds and shares an orbit lasting about 7.47 hours with a low-mass stellar companion. A. V. Karpova and colleagues report the discovery and accompanying optical observations in an arXiv preprint posted on 5 November 2026.

Redbacks belong to the broader family of “spider” pulsar systems. They pair a millisecond pulsar with a companion that is affected by the pulsar’s energetic wind. Material stripped from the companion can obscure the radio signal during part of the orbit. Unlike black widow systems, whose companions are much lighter, redbacks retain stellar companions typically weighing between roughly one-tenth and one solar mass.

The target had already attracted attention as a likely counterpart to the gamma-ray source 4FGL J2054.2+6904. Earlier optical observations showed a repeating brightness pattern near 7.5 hours, while X-ray measurements added to the evidence for a redback. Detecting radio pulses supplied the missing confirmation.

What the radio observations establish

The team observed the system with the Green Bank Telescope in four sessions between October 2025 and January 2026, collecting 7.5 hours of data. Searches accounted for the changing apparent pulse frequency caused by the neutron star’s orbital motion.

The measured radio orbital period closely matches the earlier optical period. That agreement establishes that the pulsar and the variable optical source belong to the same system, rather than being unrelated objects within the telescope’s field of view.

The observations also provide evidence for an extended radio eclipse. During the fourth session, the pulses became too faint to detect, likely because of obscuring material in the binary. Some sessions showed possible scintillation, a propagation-related variation in radio brightness.

The timing remains incomplete, however. The researchers could not uniquely count every neutron-star rotation between observing sessions: many possible timing solutions fitted almost equally well. They therefore allowed independent pulse-phase offsets between sessions. The reported orbital measurements do not amount to a fully connected, long-term timing solution.

A companion that changes unevenly

Optical observations from telescopes in Mexico and Uzbekistan sampled the companion in 2023, 2024 and 2025. They revealed a strongly asymmetric light curve—the pattern of brightness through an orbit—that simpler archival measurements had not resolved clearly.

One roughly half-orbit segment stayed relatively stable across the observed epochs. The other brightened substantially from 2023 through 2025. Because the observations came in three separated epochs, they establish long-term changes but cannot rule out additional variability over weeks or months.

Spectra obtained with the Gran Telescopio Canarias tracked the companion’s motion through shifts in its absorption lines. The team measured a radial-velocity semi-amplitude of 337 ± 4 kilometres per second. Combined with the radio orbit, this gives a pulsar-to-companion mass ratio of 7.05 ± 0.08. That ratio constrains the relative masses; determining the individual masses also requires the orbital inclination.

A potentially heavy pulsar, with model-dependent limits

To estimate the inclination and individual masses, the researchers fitted only the stable portions of the optical light curves. Their model assumes symmetric heating by the pulsar and blackbody emission from the companion’s surface.

The best fit suggests a neutron star of about 2.1 solar masses and a companion of about 0.3 solar masses. It also gives a night-side temperature near 4,960 kelvin and a maximum day-side temperature near 5,550 kelvin. The changing spectral appearance broadly agrees with a cooler face near minimum brightness and a warmer face near maximum.

These results require caution. Even the selected light-curve segments are not described well statistically: the fit has a chi-squared of 469 for 149 degrees of freedom. Additional variability is not captured by the model, so the potentially high pulsar mass should not be treated as settled.

The authors favour the dissipation or migration of a large cool stellar spot as an explanation for the changing asymmetry. This is an interpretation, not a directly observed spot. Further optical monitoring and more complete radio timing are needed to test the explanation and tighten the system’s geometry. A firmer high mass could help constrain descriptions of matter inside neutron stars; this preprint does not itself establish such a constraint.