Timing

Calibrated:
1 μsec

Uncalibrated:
~0 usec

(Absolute, not relative to radio)

PSR B1821-24 60 μsec

rxte_time.html

Timing Budget
Jahoda et al. 2006 (10.1086/500659)

PSR B1821
Rots et al. 1998
(10.1086/305836)

Crab
Rots et al. 2004
(10.1086/420842)

Event by event has 7.6 μsec resolution.

"the HEXTE absolute time reference is accurate within a fraction of a millisecond.(10.1086/305377)"

Note: Precision relative to RXTE.

Due to a wiring problem, photon time tag gets attached
to next event; correctable
under special mode for
HRC-S which telemeters
all events and then
reassigns times on the
ground. In order to avoid telemetry saturation, a higher value of the lower-level discriminator is set, which causes low PHA events to be discarded on board. This results in significant loss in QE for epochs when gain is low.

Davis et al. 2003
(davis.pdf)

Rots 2006
(CXOClock.pdf)

-354±11 μsec

Note: Timing = -306 +/- 16 μsec, Burst  = -387+/-13 μsec.
Timing mode is affected by pile up.

Note: XMM-Newton EPIC-MOS was deleted from the table.

Kirsch et al, SPIE, 5165, 85
(10.1117/12.503559)

Martin-Carrillo et al, A&A, 545, A126 (2012)
(10.1051/0004-6361/201116576)

CAL-TN-0220-1-4.pdf (Limited access)

CAL-TN-0220-1-5.pdf

L.Kuiper 2003
(10.1051/0004-6361:20031353)

L.Kuiper 2019
(13th IACHEC PDF)

Uncalibrated: tens of seconds

Calibrated:
~50 μs

BAT instrument has 100 usec time-stamping quantization, which is dominant term in uncertainty.

Offset is found by comparing RXTE PCA and Swift BAT observations of Crab (Cusumano et al 2012)

Clock measurements made by mission operations & ground station, and applied as calibration when barycentering.

G. Cusumano et al 2012
(10.1051/0004-6361/201219968)

D. Burrows et al. 2005
(10.1007/s11214-005-5097-2)

Suzaku/XIS

1.9 x 10^-9 s/s
(measured)

~ 70 μsec

360±150 μsec (90%)

(270±130  μsec 
in condition)

Freely drifting by ~3 ms/day.
Reduced to 20 us/day using
the clock correction file

Crab
B1821-24A
B1937+21

~5ms offset using
millisecond pulsars
(Lucien Kuiper)

Corrected through
clock correction file
v108+

100-300 keV band; data from Nov-2012
up to incl. Jan-2018;
Jodrell Bank radio eph. folding

316 ±70µs

D. Bhattacharya 2017 (10.1007/s12036-017-9461-x)

Basu et al (in prep)

Crab Pulsar
(with Radio) 

GRBs

D. Bhattacharya 2017 (10.1007/s12036-017-9461-x)

Basu et al 2018
(10.1051/0004-6361/201832913)

Basu et al (in prep)

Youli Tuo et al 2021

(https://arxiv.org/abs/2109.04709)

0.01μsec(GPSR)
<3.0 μsec(SpW)
0.3 ns (orbit)

~ 120-230 μsec

Uncalibrated:
1 second

Calibrated:
Approx <3 us based on Crab

Calibrated:
Measured <400 ns based on ground cal

(Absolute, not relative to radio)

The NICER XTI timing system was calibrated on ground using an end-to-end system, with GPS simulator and precision X-ray pulse generator.  Detector + system time delays of ~400 ns are included in the calibration.

Combined error / noise terms designed to be < 100 ns (including GPS, X-ray detection, barycentering)

Absolute timing offset values depend on the definition of absolute timing.
See supplementary material of this paper(https://science.sciencemag.org/content/372/6538/187),
Figure S12 and S13 [p.33-34] for
the NICER Crab X-ray main peak monitoring.
At this stage, it is not clear whether
the fluctuation of the peak phase is
due to the instrumental effects
or due to intrinsic of the pulsar.

Gulab Dewangan, Dipankar Bhattacharya