One of those long, quiet slews that Exosat used to carry out to point its telescopes and detectors from one target to the next one in the sky was on for that evening of May 18th, 1985. Long slews were a boring operation, especially after having done so many in two years of mission. The spacecraft controller on shift monitored the status of the on-board subsystems and the correct evolution of the attitude angles. The X-ray telescope was switched off, as foreseen during all slews, while the other, less delicate X-ray detectors were kept active to monitor the sky background and to observe possible new sources. The sky observed in the X-ray frequencies is very turbulent. New sources (called “transients”) can suddenly appear and disappear within a few hours, as result of catastrophic stellar events or as part of the normal life of celestial objects at distances of thousands of light years from us.

I was in the room near the Dedicated Controlled Room (DCR), in which the HP1000 computers for the scientific data analysis and the access terminals were located. One of the observatory scientists, Arvind Parmar, was in the same room. It was 18:54 GMT, that is about 9 p.m., when we heard the alarm ringing, and the spacecraft controller called us from the DCR: the Medium Energy detector had reached the alarm threshold on the number of counts. I produced a graphic plot on the workstation, which showed a narrow peak of the detected counts. This could only mean one thing: a source. Since no known source of this intensity was predicted to pass over our detectors during the slew, we were already quite sure that Exosat had detected a new X-ray star! Arvind rushed in the computers room, quickly transferred the telemetry data to the science computers and produced, within minutes, a power spectrum of the signal. He came back in the DCR with a smile and a hardcopy in his hands. He showed the spectrum to us. The shape was a power law, the typical signature of a X-ray pulsar. We had some software tools to determine with enough accuracy the position in the sky at which Exosat was pointing the instruments at the time of the event. After having calculated the position, we passed the information to our mission planning team to obtain some high priority observation time for the new target. The Exosat observatory had the flexibility to re-arrange observation slots and targets on short notice, if important scientific reasons required it. And a new pulsar was important enough. In fact we carried out this first observation on the new source in the same night at 22:43, just a few hours after the discovery.

In the following days we carried out several observations of the pulsar, which was named very poetically EXO 2030+375, after its position in the celestial sphere. A X-ray pulsar is a very small (of the order of a few km diameter), extremely dense and quickly rotating remnant of a dead star, which orbits around another star, from which it attracts plasma (ionised gas) due to its intense gravitational field. The plasma is accelerated and emits high energy radiation, which is observed by X-ray detectors above the Earth atmosphere. If the orbit of the pulsar around the companion star is very elongated, the pulsar only “sucks” plasma from the star when it comes close to it, and therefore it only emits X-rays in limited parts of its revolution. This is why we have “transient” X-ray pulsars, and this is why it is so difficult to discover them.

The light curve, that is the intensity of the X-ray signal over the rotation period (42 seconds) was determined. It was a very typical, stable light curve that could be observed in other X-ray binary systems. The small variations of the pulsar rotation period measured in the various observations allowed us to estimate the period of revolution of the pulsar around its companion star (between 44 and 48 days). We could now reasonably well predict the next time in which our transient X-ray pulsar would become visible again. Of course we planned new observations with Exosat around the predicted time of maximum intensity.

The day of the start of the new observation campaign, on October 29th, at 15:59 GMT, I happened to be again on shift in the control room. Peter Vogt, the spacecraft controller on shift, announced that the slew to the target was completed, and I could see on my screen the rising signal of EXO 2030+375. This was a typical effect when a bright source was coming in the field of view of the detectors. In this case, though, the signal continued to rise even minutes after the end of the slew. I asked Peter: “When do you stabilise the spacecraft, Peter?”. He replied, calmly: “It is stable like a rock”. He showed me the readings of the attitude control sensors, confirming the pointing attitude was stable within the nominal limits. Something did not fit: If the pointing was stable, why was the X-ray signal rising? Nick White, the chief observatory scientist, Arvind Parmar and Luigi Stella were there, scratching their heads. There was only one explanation: We were witnessing a X-ray flare, that is a sudden variation of the signal intensity of the X-ray pulsar. This was a behaviour that we had not observed at all in the first campaign, and it made the source even more exotic and exciting! The light curve had also changed shape, and it kept changing its shape during the entire second observations campaign. In the following observation we kept Exosat pointing at the source for 23 hours, observing six flares each lasting 1.3 to 2.3 hours and repeating every 3.96 hours. It was fantastic to see this object – “our” pulsar – behaving in such a beautiful and unexpected way! The fact that the periodicity of the flares could not be related to the orbital period was observed for the first time in a X-ray binary pulsar.