How do I know whether my HRI data is affected by one of the subtle timing problems I've heard about?
(3/1/94)
HRI Timing Summary
All HRI data processed at GSFC after Oct 22nd 1992 (when HRI4_5_2 was installed in the on-line processing) are free of timing problems.
Background
The Space-craft clock onboard the ROSAT satellite has undergone several resets during the first two years of the mission. There have been two reasons for space-clock resets:
- the satellite is lost or enters safe mode
- annually due to an overflow in the counter
Until recently, the processing of these data has introduced an fractional second offset in times assigned to the photons. Below is a list of the clock resets and the fractional seconds by which the times are offset.
1991:
- 26 Jan
- Satellite Glitch. HRI data taken after this time have event
times with an absolute offset of 33/64 seconds.
- 25 June
- Annual overflow. HRI data taken after this time have event
times with an absolute offset of 58/64 seconds.
1992:
- 12 Feb
- Satellite Glitch. The reset of 63/64 second caused the most sever
problems in the HRI software. Data taken after this time can have
backward times of less than one second in duration. Timing analyses
for which timescales of a few seconds or less are important may be
affected and should be deferred until reprocessed data become
available. Fixed in Release SASS5_8 (HRI_4_5).
- 17 July
- Annual Overflow. Negative fractional reset of -1/64 second.
The event times have an absolute offset of -1.0 second.
This absolute offset should not inhibit users from searching
for periodic behavior. Fixed in SASS HRI4_5_2.
- If an Observation Interval (OBI) run previous to HRI 4_5 is merged with one run post HRI 4_5, the times in the early OBIs will be offset by a constant fractional second, and the new OBIs will be correct to the fraction. All data run with a software version previous to HRI 4_5 and since the first reset will also be offset by the fractional seconds of the reset. This value will be constant throughout the sequence unless OBIs that were merged were taken at different resets.
- If an Observation Interval (OBI) run previous to HRI 4_5 is merged with one run post HRI 4_5, the times in the early OBIs will be offset by a constant fractional second, and the new OBIs will be correct to the fraction. All data run with a software version previous to HRI 4_5 and since the first reset will also be offset by the fractional seconds of the reset. This value will be constant throughout the sequence unless OBIs that were merged were taken at different resets.
Select another question related to this topic
Select another FAQ topic Return to the ROSAT GOF
Please use the Feedback link if you have questions on ROSAT.