Large defect was developed in chip-2 of SIS-1 between May 14 and 25 in 1999. The defect may be located in the frame-store region. Because the leak current from the defect fills most of the pixels in the chip, the chip became useless for scientific observations. The cause of the development of the defect is not known.
Frame data taken to investigate the nature of the defect are shown below in chronological order.
| Date (UT) | Caption | Image |
|---|---|---|
| 1999 May 14, 16:39 | The last frame data (16 sec exposure) obtained when the chip was healthy. The data were normal and we could not find any hint of developing a big defect. A dark column in the lower right of the image is due to the presence of a small defect pixel. Such columns are sometimes seen in the SIS chips and are harmless for the X-ray observations. |
|
| 1999 May 25, 21:25 | The first frame data (16 sec exposure) taken after the appearance of the defect. Yellow color corresponds to 4095 adu, which means the data were saturated. Note that first 42 rows seem to be normal. |
|
| 1999 May 27, 11:03 | Frame data (4 sec exposure) acquired when the back-bias was changed to the low level. Because no clear change was noticed in the frame data, back-bias is not related to the leak current. |
|
| 1999 May 28, 4:10 | Frame data (4 sec exposure) acquired when the AE gain was set low (1/4 of the high gain). The unsaturated region beame a little wider, but still most of the pixels were saturated. Note that the v-overclocked region was also saturated as the imaging region, whereas the first 42 rows were almost free from the leak current. This may be understood if the defect was developed in the frame-store region. |
|
| 1999 May 28, 9:10 | AE was set in the diagnostic mode. In this mode, CCD is clocked for the reverse direction. Thus no net charge is transfered to the read-out gate. The image was same as that obtained in the ground test and verified that the on-chip FET and AE were functioning correctly. Interference pattern, which was also observed during the ground test, was recognized in the magnified image. |
|
| 1999 May 29, 3:55 | Frame data (4 sec exposure) acquired when the imaging area high/low clock voltages were both set to zero. In this setting, no charge in the imaging area was transfered to the frame-store region. Thus, this experiment clearly showed that the defect was produced in the frame-store region. |
|
| 2000 June 2, 1:44 | Frame data (16 sec exposure) acquired about a year after the emergence of the defect. There seems to be little change in the leak current from the defect. |
|
From the various experiments we carried out, we could obtain following characteristics on the defect:
Back