PACS Photometer glitch analysis on the electronics Page 1 PACS Photometer glitch analysis on the electronics K. Okumura
PACS Photometer glitch analysis on the electronics Page 2 Req. 1.1.9bis Glitches on electronics components 1.1.9bis - A. History Version Date Author(s) Change description 1.0 13 SEP 2010 K.Okumura First issue 1.1.9bis - B. Summary At the beginning of the mission up to OD 160, we used the stabilisation time after switch on to measure the reference voltage VRL. Those measurements are affected by the slow stabilisation of the signal, but some usefull information can still be found in those data. Here we use those data to asses the glitch behaviour on electronics. 1.1.9bis - C. Data Reference Sheet Table 1: Data used for the analyses OD date OBSID comment 026 09 Jun 2009 1342178014 064 17 Jul 2009 1342180037 065 18 Jul 2009 1342180117 067 20 Jul 2009 1342180136 072 25 Jul 2009 1342180669 073 26 Jul 2009 1342180761 086 08 Aug 2009 1342181644 092 14 Aug 2009 1342182091 093 15 Aug 2009 1342182234 096 18 Aug 2009 1342182403 101 23 Aug 2009 1342182610 104 26 Aug 2009 1342182822 107 29 Aug 2009 1342182942 110 01 Sep 2009 1342183024 120 11 Sep 2009 1342183649 124 15 Sep 2009 1342183877 128 19 Sep 2009 1342183989 no data 132 23 Sep 2009 1342184263 139 30 Sep 2009 1342184570 157 18 Oct 2009 1342186088 160 21 Oct 2009 1342186131 1.1.9bis - D. Test Description Right after the switch on of the PACS photometer, the detectors need about 35 minutes to stabilise. We take advantage of this stabilisation time to measure the signal coming out of the reference voltage VRL. This signal does not include the bolometer. However, the signal can be affected by the cosmic ray impacts on the electronics components and can contain some electronics specific glitches.
PACS Photometer glitch analysis on the electronics Page 3 1.1.9bis - E. Results In the signal from the reference voltage VRL there is no glitch comparable to the short time constant glitches seen on the bolometer signal (Fig. 1). However, glitches of long term recovery can be detected with a much less occurrence. Typical examples of those long therm glitches are shown in Figs. 2, where one observes negative glitches as well as positive ones. Figure 1: A typical example of the VRL signal after the bolometer switch on and during the stabilisation. No usual glitch is seen. In order to detect this type of signal behaviour the usual deglitching algorithm does not work in a optimum way. Here the detection is performed by a fixed threshold of the absolute value of the derivative of the signal. The threshold was fixed here to 10 times the standard deviation around the median value of the derivative of the signal. The number of detected glitches using this method is on average 33 over roughly 30 minutes on the whole blue detector area. Taking this small amount of glitches into account, one can construct a global view of the glitches over this time interval by looking at the largest glitch by pixel. A map of the time of impact can be constructed by mapping the frame number of the largest glitch for each pixel. This map allows to see the coincidence of the glitch on several pixels and their spatial distribution. Another map can also be constructed to show the amplitude of the largest glitch detected on each pixel, which allows to see, in particular, if the glitches are positive or negative. Figs. 3 to Figs. 22 show those maps for all the available data from OD26 to OD160. One thing is clearly noticeable in these maps: the pixels on the 11th line of the matrix 6 (upper inner left of the blue array) are almost always detected as glitches. This systematic detection suggest that this is not due to cosmic ray hits, but rather to the electric noise of this readout address.
PACS Photometer glitch analysis on the electronics Page 4 Figure 2: Example of glitches. [Top] typical positive and negative glitches, [Middle] glitches of a relatively large amplitude, [Bottom left] successive glitches (the first one affects only this pixel whereas the second one is seen on 16 pixels of the same readout address, [Bottom right] this should be considered rather as electric noise (see hereafter)
PACS Photometer glitch analysis on the electronics Page 5 Figure 3: Global view of the glitches OD 26. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red Figure 4: Global view of the glitches on OD 64. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red
PACS Photometer glitch analysis on the electronics Page 6 Figure 5: Global view of the glitches on OD 65. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red Figure 6: Global view of the glitches on OD 67. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch. The result is only on the blue side, for no glitch was detected on the red side.
PACS Photometer glitch analysis on the electronics Page 7 Figure 7: Global view of the glitches on OD 72. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch [Top] blue, [Bottom] red Figure 8: Global view of the glitches on OD 73. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch. The result is only on the blue side, for no glitch was detected on the red side.
PACS Photometer glitch analysis on the electronics Page 8 Figure 9: Global view of the glitches on OD 86. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red Figure 10: Global view of the glitches on OD 92. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red
PACS Photometer glitch analysis on the electronics Page 9 Figure 11: Global view of the glitches on OD 93. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red Figure 12: Global view of the glitches on OD 96. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red
PACS Photometer glitch analysis on the electronics Page 10 Figure 13: Global view of the glitches on OD 101. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red Figure 14: Global view of the glitches on OD 104. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red
PACS Photometer glitch analysis on the electronics Page 11 Figure 15: Global view of the glitches on OD 107. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch. The result is only on the blue side, for no glitch was detected on the red side. Figure 16: Global view of the glitches on OD 110. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red
PACS Photometer glitch analysis on the electronics Page 12 Figure 17: Global view of the glitches on OD 120. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red Figure 18: Global view of the glitches on OD 124. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red
PACS Photometer glitch analysis on the electronics Page 13 Figure 19: Global view of the glitches on OD 132. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red Figure 20: Global view of the glitches on OD 139. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red
PACS Photometer glitch analysis on the electronics Page 14 Figure 21: Global view of the glitches on OD 157. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red Figure 22: Global view of the glitches on OD 160. [Left] the frame sequence number where the largest glitch is detected, [Right] the amplitude in volts of the corresponding glitch, [Top] blue, [Bottom] red
PACS Photometer glitch analysis on the electronics Page 15 In fact a typical signal of one of these pixel looks like in Fig. 23, where the signal level skips suddenly on slightly a lower level from time to time and then skips back to the previous level. If we look at the average value of the number of detected glitches, it is 40 for the blue and 7 for the red. The geometrical retio of detector surface being 4, the glitch detection is higher for the blue than the red. If we subtract all the detection on the 11 th line of the matrix 6, we obtain 29 instead of 40 and the ratio 29/7 is nearly 4. Figure 23: The pixels in the 11th line of the matrix 6 are always detected as glitches. These detection are too systematic to consider them as due to cosmic ray hits. It should be considered as electric noise due to a instable electronics on this address. Now taking into account this false detection, we can remove all the detection on the line 11 of the matrix 6, without introducing too much bias because of the small number of glitches. Figs. 24 show the number of detected glitches corrected by the electric noise. The fluctuation of the glitches in time does not correlate between blue and red detectors. Coming back to the glitch maps, some features are easily noticeable. There are glitches affecting only one pixel There are glitches affecting a set of pixels of the same readout address There are as many positive as negative glitches in both cases Figs. 25 to Figs. 34 show a set of glitches occurred on a same readout address. In the data of OD26, the pixel [28,3] shows an noise similar to that of the line 11 of the matrix 6. The same kind of noise can be also seen on the line 31 of the matrix 1 on OD160.
PACS Photometer glitch analysis on the electronics Page 16 Figure 24: Number of detected glitches. [Left] Blue and [Right] Red Ideally the statistics of the glitch rate of those different features are supposed to reflect the geometrical cross section of electric componets. This correspondance should be studied further in detail. 1.1.9bis - F. Conclusions The observed glitches on the reference signal VRL consist of two groups. The ones which affect symaltaniously the pixels of one readout address and those which affect only one independent pixel at the same time. 1.1.9bis - G. IA scripts used / remarks on PCSS glitchesvrl.py glitchesvrlstat.py
PACS Photometer glitch analysis on the electronics Page 17 Figure 25: Glitches of a same readout address on the matrix 1
PACS Photometer glitch analysis on the electronics Page 18 Figure 26: Glitches of a same readout address on the matrix 2
PACS Photometer glitch analysis on the electronics Page 19 Figure 27: Glitches of a same readout address on the matrix 3 Figure 28: Glitches of a same readout address on the matrix 4
PACS Photometer glitch analysis on the electronics Page 20 Figure 29: Glitches of a same readout address on the matrix 5
PACS Photometer glitch analysis on the electronics Page 21 Figure 30: Glitches of a same readout address on the matrix 6
PACS Photometer glitch analysis on the electronics Page 22 Figure 31: Glitches of a same readout address on the matrix 7
PACS Photometer glitch analysis on the electronics Page 23 Figure 32: Glitches of a same readout address on the matrix 8
PACS Photometer glitch analysis on the electronics Page 24 Figure 33: Glitches of a same readout address on the red matrix 9
PACS Photometer glitch analysis on the electronics Page 25 Figure 34: Glitches of a same readout address on the red matrix 10