GIS Status Report: January 24, 1994
--edited by K. Makishima
Status of the DetectorsThe GIS detectors have been functioning normally since the launch. Although the gain dropped by 2 percent and 3 percent respectively for the GIS-2 and -3, during the first 5 months, the gain now gradually is attaining a constant level. The gain degradation rate now is almost negligible. This level of gain drop was foreseen before the launch and will cause no damage to observations for several more years.
During the Crab calibration (early April in 1993), almost all of the GIS control parameters were fixed. The high voltages (HVH/L = -6000/+1120V) and the fine gains were adjusted so that the energy range of 0-11keV is covered. RT windows are set to safely accept signal X-rays and to significantly reject the background. The algorithm of event position calculation has been fixed to FLF (Fast Lorentzian Fit). The only control paramter not fixed by April 1993 is the SPREAD discriminator window; which was fixed on May 27, 1993. After that, the background rate was reduced by 40% in full area. SPREAD discriminator is more effective at the edge of the detector. Observers analyzing extended sources thus should be careful in choosing background data.
Response ProblemsThe latest version of the RMF response function and ARF builder are gis*v2_4.rmf and jbldarf-0.45 (where * is 2 or 3 representing GIS-2 or -3), respectively. We have been evaluating the quality of these responses with the data of Crab, Cyg X-1, and Cyg X-2 since October '93.
In fitting this combination of responses to the Crab data, we obtained the photon index of 2.06-2.10 and the column density of (0.36-0.40)x10^22, depending upon the source position and/or the sensor ID. The following two major problems still remain in the response function:
- The measured column density is somewhat higher than the values
indicated by previous observations, which are typically
- Although the photon index is consistent with that obtained by previous
missions, data show significant excess above the model, which amounts
to 20-30% at 10~keV. (Note that a similar problem was found in the course
of BBXRT calibration.)
Evaluation of O-tailThe distribution of event pulse heights for monochromatic X-rays obeys a so-called O-distribution, namely, an approximately Gaussian core with a tail extending to lower pulse heights. The tail is generated due to absorption of a part of electrons by the beryllium entrance window. The magnitude of the tail is determined by a competition between the drift velocity of the electron cloud and its diffusive expansion. Since lower energy X-rays are absorbed closer to the Beryllium window, the tail is more conspicuous for lower energy X-rays. Presently, we know that the amount of O-tail is overestimated by about 30 percent, which will enhance the efficiency in the lower energies and may well result in the apparently larger hydrogen column density for the Crab. This issue will be corrected in the next version of the RMF response.
Boresight positionBoresight positions of the GIS-2 and -3 are revised based on the second Crab calibration. They are as follows:
- (+1.00mm, +0.49mm) for GIS-2,
- (-2.41mm, +1.36mm) for GIS-3.
Gain map and absolute gainThrough a detailed comparison using W49B, we found no significant difference in the effect of gain equalization (as a function of position) between production software and QL software. However, it seems that the absolute gain of GIS-3 is higher than that of GIS-2 by approximately 1 percent. The center energy of iron K-line from W49B is now at 6.65keV.
It was a matter of question at what energy the calibration source should appear after gain equalization using the current version of calibration data. According to Ikebe-san, the energy should be as follows:
- 5.82-5.87 keV (494-498 ch) for GIS-2
- 5.56-5.61 keV (472-476 ch) for GIS-3
Grid MapUsing daytime earth data and some cluster data, Matsushita-san has been evaluating the effect of the grids. Although there still remain some uncertainties for S2, the grid map of S3 is mostly fixed. Results obtained will be reflected in the next version of the grid map.
It has been very difficult to evaluate the efficiency decrement when a point source is right on the grid. This difficulty is because such data were taken neither in ground-calibration phase nor in the PV phase. There are some data during GO phase. Matsushita-san is evaluating the effect of the grid for the point source with these data. This point should be settled down by middle February.
Beryllium thickness map
Using coma cluster data, thickness of the beryllium is evaluated. Ueda-san has confirmed a gradient of thickness for S2 that is indicated by the ground calibration. The thickness is uniform (=10.7um) for a region DETY < 0, and monotonically increased by increasing DETY from 0. The thickness becomes 11.1um at DETY = +20mm. On the other hand, the uniformity of the S3 beryllium is very good and the thickness is within 0.1um in full amplitude. (As a reference, 0.2um of beryllium is roughly equivalent to 1x10^20 of hydrogen column density at about 1 keV assuming solar abundance.) These results will be taken into account in the next version of the beryllium thickness map.
ScheduleIn reference to the response problems previously listed, the XRT team and the GIS team agree that the column density should be adjusted by the GIS team, while the photon index should be adjusted by the XRT team. The items related to the column density problem of the response function are the evaluation of the O-tail and the Beryllium thickness map. We hope to complete actions on these two items by early February. However, it is difficult to give conclusions on the gain map absolute gain and the grid map until mid-February.
Dead Time CorrectionsThe dead time of the GIS is calculated by ADC dead time, the number of X-ray events appearing in the telemetry, and several other monitor counting rates. All the words required now are output by MKFILTER ver1.23. A model constructed by these quantities can reproduce the dead time correctly.
Ishisaki-san evaluated the counting rates that make the dead time 10%. They are as follows:
- Bit H 60 c/s ( CPU limit)
- Bit M 16 c/s ( telemetry limit )
- Bit L 4 c/s ( telemetry limit ).
Below these counting rates, the deadtime gradually becomes negligible.
BackgroundBy the analysis of night earth data during the PV phase, it is revealed that the non-X-ray background spectrum is possibly reconstructed by the position of the satellite on the earth and the two monitor words called L2 and H2. A program for constructing non-X-ray background based on this idea has been considered.
We are considering the possibility of supplying all the ASCA users with a sample CXB spectrum.
It should be discussed among the ASCA team members whether it is possible to open photon files of blank skies taken during the GO phase.
For more information regarding this document, contact:
- Manabu ISHIDA
- Institute of Space and Astronautical Science (ISAS)
- e-mail: firstname.lastname@example.org
- Phone: 81-427-51-3911 ext.2612
- FAX: 81-427-59-4253
- e-mail: email@example.com
If you have any questions concerning ASCA, visit our Feedback form.